Help
Options
Didn't know it?
click below
click below
Knew it?
click below
click below
Don't Know
Remaining cards (0)
Know
retry
shuffle
restart
0:00
Embed Code - If you would like this activity on your web page, copy the script below and paste it into your web page.
Normal Size Small Size show me how
Normal Size Small Size show me how
Stack #2549802
| Question | Answer |
|---|---|
| There are 2 types of statitiscs (Analytics) | descriptive & inferential |
| descriptive statistics are used to | inform/explantory |
| inferiental statistics are used to | predict/trend |
| name 4 levels of measurement | (NOIR) Nominal, ordinal, interval, ratio |
| continuus data w/unique zero point | ratio |
| orders data at EQUAL distance apart | interval |
| place qualitative objects in some kind of order | ordinal (order) |
| indentify, group or catergorize | nominal (neat) |
| outliers create this type of error | out-of-range |
| unpredictable error | random error - no correlation |
| error may occur from missing data. (example: space not filled in) | omission error - distorted results |
| this error repeats itself | systematic error - skewed results |
| what is the purpose of the process of quality control | reduce/minimize errors |
| experimental study | all variable measurements & manipulations are under the researcher's control |
| observational study | used with impractical or impossbile to control the conditions of the study |
| blind study | participants are not told if they are in the treatment gorup or control group |
| treatments | the procedure the researcher applies to each subject |
| double blind study | neither the treatment allocator or the participants knows who is in the treatment group or control group |
| information bias | questions favor an outcome or the interviewer asks questions that favor an outcome |
| expected monetary value (EMV) anaylsis | the average outcome (payoff) when the future includes scenarios that may or may not happen |
| outliers | observation points that are distrant from other observations |
| measurement bias | bias that occurs from not selecting a random sample |
| conscious bias | bias introducted because respondents believe it will be beneficial if selected |
| median - outliers do not affect the median | middle score for a data set |
| z-score | tells us the number of standard deviations a data point is from the mean |
| variance | if the average is the same for 2 groups, what will determine their difference? |
| standard deviation | the spread of data in a sample. |
| standard deviation | how far the data points are from the mean |
| mean | measure of central tendency that is influenced by the size of the values in a dataset |
| quartiles | each of the four quartile groups a population can be divided into |
| IQR: Inter-quartile range | must be ordered in lowest to highest value |
| box plot | used to study the composition of a data set & determine the distribution |
| precentile score order: | max value, 25th percentile, median, 75th percentile, min value |
| mean & median | skewedness does not affect these |
| outliers | Can be included or excluded in anaylsis (causes skewness) |
| 1 customer & 6 booths = 1/6 or 16.7% | There are 6 toll booths to enter the highway. What probability does ecah toll booth worker have of getting the next customer |
| combination | the order you pick your sample in does not matter |
| combination | Example: picking employees for a shift. Order doesn't matter |
| Bayes Theorem | you must know P(A), P(B) |
| Bayes Theorem | P(A) given B |
| Bayes Theorem | When given P(A) given P(B), you can use this to find the P(B) given P(A) |
| multiplication | apply this rule when looking for 2 events occuring (AND) |
| addition | use this rule when looking for 1 of the other event happening (OR) |
| linear programming | a technique for minimizing total cost or maximizing profit based on constraints |
| linear regression | a technique using a single independent variable to predict a single dependent variable |
| multiple regression | a technique using more than 1 independent variable to predict a single dependent variable |
| correlation coefficient | measures the strenth of a linear relationship |
| R2 or R squared | measures the goodness of fit in a regression anaylsis |
| statistical tests | R2, correlation coefficient, mutliple regression, linear regression, linear programming |
| descriptive statistic measurements | median, mean, z-score, variance, standard deviation |
| types of bias | information, outliers, expected monetary value (EMV), measurement, conscious |
| types of studies | experimental, observational, blind, treatments, double blind |
| types of errors | out-of-range, random, omission, systematic |
| levels of measurement | ratio, interval, ordinal, nominal |
| distribution & central tendency | time series, trend, irregularity, cyclicality, seasonality |
| time series | a simple regression using time as the independent variable |
| trend | a general slope upward or downward over a period of time |
| irregularity | unforseen circumsntances causing random deviations |
| cyclicality | reptition in up and donw patterns |
| seasonality | regular pattern within a single year |
| negative | from left to right: arrow slating from top to bottom (downward) |
| positive | from left to right: arrow slanting from bottom to top (upward) |
| strong | dots are close to together |
| weak | dots are spread apart |
| cumulative distribution | represents the probability tha a variable falls within a certain range |
| probbility distribution | a list of all fo the different probabiities of each outcome that can occur |
| z-score for 99% level of confidence | 2.575 |
| z-score for 95% level fo confidence | 1.96 |
| normal distribution | measures of central tendance are approximately equal (mean & median) |
| ANOVA | used to compare the mean of three or more groups |
| anaylsis techniques | ANOVA, trend, irregulairty, cyclicality, seasonality |
| F-value | must be higher than critical value to reject the null |
| F-value | ANOVA uses this test statistics |
| T-value | must be higher than critical value to reject the null |
| T-value | T-test uses this test statistic |
| zero | a correlation is weak if the coefficient is close to this |
| 1 or -1 | a correlation is strong if the coefficient is close to this |
| 7 basic quality tools | run chart, control chart, case & effect diagram, flowchart, check sheet, histogram & pareto, scatter diagram |
| run chart | illustrates peformance measurements over a period of time |
| control chart | illustrates limits or constratins a process should not exceed |
| cause & effect diaggram | assists in brainstorming issues that are causign a problem |
| flowchart | visual tool to understand a process |
| check sheet | easy tool to collect data to create other charts |
| histogram & pareto | graphical display of a data set with 1 bar for each cateory |
| pareto | graphical display of data set in highest to lowest order |
| histogram | graphical display of data set centered |
| scatter diagram | used for potential relationships & correlation between vairables |
| yes | can the seven basic quality tools be used independently |
| 90%-95% | what percent of quality problems does Ishikawa claim the 7 tools can solve |
| SIPOC | supplier-input-process-output-customer |
| SIPOC | diagram demonstrating all of the elements that can influence a process before it starts |
| six sigma | manufacturing approach to improving processes |
| quality control | in manufacturatin, statistics are used for this purpose |
| Act | the step in the PDSA that is a response to analytical results |
| PDSA | Plan-Do-Study-Act formerly PDCA (Plan-Do-Check-Act) |
| attribute | shows whether a result meets a requirment or not |
| variable | shows how well a result meets the requirement |
| common cause variation | variations accepted as a normal part of the process |
| special cause variation | variation from an abnormality causing large discrepancy in results |
| IRT | Item response theory |
| IRT | model of designing, anyalzing & scoring tests |
| goventment benefits aren't always money | could be flood prevention or welfare |
| how does government differ from private sector cost-benefit analsysis | government beneifts aren't always money |
| norm referenced | compares one individual's performance to other individuals |
| criterion referenced | compare individual's performance to a standard score (ex: cut score 64%) |
| cost-benefit analysis | used to analyze if funding is worth the outcome of a project |
| very large data sets | what is big data |
| prevalence | used to count all of the exisint cases in a disease |
| incidence (incident rate) | used to count only NEW cases of a disease |
| used in big data | very large data sets, prevalence, incidence, criterion referenced, cost-benefit anaylsis |
| RBM | Results based management |
| RBM | management strategy that uses results as the central measurement of performance |
| performance measures | KPI, KPI dashboard, CLIF |
| KPI | Key performance indicators |
| KPI | performance measure for 1 specific goal |
| KPI dashboard | multiple KPIs are displayed for the big picture |
| CLIF | customer learning, internal process, financial performance |
| What does a balance scorecard measure | CLIF |
| What does a balance scorecard measure | Are we meeting the strategy? |
| Disadvantage of balanced score card | requires time & effort to establish a meaningful scorecard |
| advantage of a balanced scorecard | improves internal & external communication |
| Disadvantage of balanced score card | dificult to maintain momentum |
| advantage of a balanced scorecard | improves organizational alignment |
| advantage of a balanced scorecard | links strategy to organizational results |
| advantage of KPI | data driven results make it easier to quantify performance |
| disadvantage of KPI | difficult to change once set up |
| garbage in garbage out | bad data produces bad results/decisions |
| analytics | the extensive use of data, statistical & quantative analysis, explanatory & predictive models & fact-based management to drive decisions & add value |
| analytics emcompases | statsitics or the study of data analysis |
| analytics emcompases | management science or the study of model building, optimiation & decision making |
| analytics | typically implies the analysis of very large data sets |
| analytics is | turning information into insight & developing conclusive fact-based startegies to gain a competitive edge |
| you can't manage | what you don't measure |
| big data refers to | both structured & unstructured data in large volumes |
| structured data | example: credit card transactions through a website |
| unstructured data | word documents or the bodies of emails |
| unstructured data | videos from a traffic accident sitte |
| unstructured data | data that doesn't reside in a traditional row-colum database format |
| managers are | constantly called upon to make decisions in order to solve problems |
| decision making & problem solving | ongoing processes of evaluationg situations or problems, considering alternatives, making choices & following them up with necessary action |
| the entire decision-making process is dependent upon | the right information being available to the right people at the right ime |
| quantitive decision making steps | framing the problem, solving the problem, communicating results |
| framing the problem includes | broad definition of business problem |
| framing the problem includes | review what has happened in the past |
| framing the problem includes | type of anaylsis |
| framing the problem includes | what data & how to collect |
| solving the problem includes | specific question to analyze |
| solving the problem includes | type of data, data collection & data error |
| solving the problem includes | analysis technique & analysis |
| communicating results includes | presentation of analysis output |
| communicating results includes | making a recommendation |
| analytics is | turning information into insight |
| value and difficulty increase when | going from information to optimization |
| value and difficulty increase when | when going from hindsight to insight to foresight |
| hindsight | descriptive analysis & diagnostic analysis |
| descriptive analysis | what happened |
| insight | predictive analysis |
| diagnostic anlaysis | why did it happen |
| foresight | prescriptive analysis |
| predictive analysis | what will happen |
| prescriptive analytics | how can we make it happen |
| descriptive/diagnostic analytics | encompasses the set of techniquest that descripe what has happened in the past |
| descriptive/diagnostic analytics | the discovery & communiation of meaningful patterns in data |
| descriptive/diagnostic analytics | the vast majority of the stats we use fall into this category |
| descriptive/diagnostic analytics | think basic arithmatic like sums, averages, percent changes |
| descriptive/diagnostic analytics | utilizes graphical analysis such as scatter grams, histograms, pareto charts & run charts |
| descriptive/diagnostic analytics | useful to show total stock in inventory, avg dollars spent per customer & year over year changes in sales |
| common exampes of descriptive/diagnostic analytics | reports that show historical data regarding a company's production, financials, operations, sales, fiannce, inventory & customers |
| predictive analysis | consists of techniques that use models constucted from past data to predict the future or ascertain the impact of 1 variable on another |
| predictive analysis | uses models such as regression, time series & statistical quality control |
| predictive analysis | uses simulation |
| predictive analysis simulation | involes the use of probablity & stats to construct a computer model to study the impact of uncertainty on a decision |
| prescriptive analytics | decision models that indicate the best course of action to take |
| prescriptive analytics | uses optimization models, simulation optimization & decision analysis |
| optimization models | models that suggest the best descision subject to constraints of the situation |
| simulation optimiation | combines the use of probablity & stats to model uncertainty w/optimization techniques to find good decisions in highly complex & highly uncertain settings |
| decision analysis | used to devleop an optimal strategy when a decision maker is faced with several decision alternatives & an uncertain set of future events |
| decision analysis | employs utlity theory, which assigns values to outcomes based on the decision maker's attitutude towards risk, loss and other factors |
| marketing | one of the fastest growing areas for the application of analytics |
| resulted in a better understanding of consumer behvior | the use of scanner data & data generated from social media |
| resulted in an increased interest in markeing analystics | a better understanding of consumer behavior obtained through the use of scanner data & data generated from social media |
| a better understanding of consumer behavior through marketing analytics leads to: | the better use of advertising budgets |
| a better understanding of consumer behavior through marketing analytics leads to: | more effective pricing strategies |
| a better understanding of consumer behavior through marketing analytics leads to: | improved forecasting of demand |
| a better understanding of consumer behavior through marketing analytics leads to: | improved product line management |
| a better understanding of consumer behavior through marketing analytics leads to: | increased customer satisfaction & loyalty |
| levels of measurement: data | breaks down into measurement and counting |
| levels of measurement: measurement | breaks down into ratio & interval |
| levels of measurement: counting | breaks down into ordinal & nominal |
| levels of measurement: ratio & inteval | are numerical & continuous |
| levels of measure: ordinal & nominal | are categorical & discrete |
| nominal level of measurment | lowest of the 4 ways to characterize data |
| nominal data | (naming) deals with names, categories or labels |
| nominal data | is qualitative |
| examples of nominal data | colors of eyes, yes or no survey response, favorite cereal |
| nominal data level | data at this level can't be ordered in a meaingful ways |
| nominal data level | makes no sense to calcualte things such as means & standard deviations |
| ordinal data | places data objects into an order according to some quality |
| ordinal data | should not be used in calcuations |
| nominal & ordinal data | should not be used in calcuations |
| ordinal data | (organizing) examples incude a listing of top 10 cities in which to live |
| nominal data | naming |
| ordinal data | organizing |
| interval data | has an order to it & all objects are an equal interval part |
| interval data | the difference between 2 values is meaningful |
| interval data | does not have a natural zero point |
| interval data | zero does not represent the absence of the property being measured |
| interval data | example: Fareheit & celsius temperature scales. You can talk about the differences in temps, 0 does not equal the absense of temperature. |
| interval data | this data can be used in calcuations |
| ratio data | similar to interval data but has a unique zero point |
| ratio data | numbers can be compared as mutliples of one another |
| ratio data level | not only can sums & differences be calculated but also ratios |
| ratio data | one measurement can be divided by any nonzero measurements resulting in a meaningful number |
| in business | ratio data is common |
| ratio data | examples include: income, stock price, amount of inventory & # of repeat customers. All have a unique zero potin & can be compared through mutliplication |
| data management | process by which the required data is acquiared, validated, stored, protected & processed |
| data management | process by wich the accessiblity, reliability & timeliness of data is ensured |
| data management | includes the cleaning, organizing & storage of collected data |
| business decisions | are not better than the data on which they are based |
| reliable, relevent and complete data | supports organizational efficiency & is a cornerstone of sound decision-making |
| complete | all relevant data - such as accounts, addresses & relationships for a given customer - is linked |
| accurate | common data problems like misspellings, typos, and random abbreviations have been cleaned up |
| available | required data is accessible on demand; users do not need to search manually for the info |
| timely | up-to-date informaiton is readily available to support decision |
| all measurements contain | some degree of error |
| random error | in experiemental measurements are cuased by unknown & unpredictable changes in the experiement |
| random error | these changes may occur in the measruing instruments or in environmental conditions |
| random error | should cancel themselves out over a large number of measurements |
| systematic error | systematic errors in experiemental observations usually come form the measuring instruments or experiemnental design |
| systematic error | measument errors that are constant within a data set |
| omission error | occurs when action has not beeen taken or when something has been left out |
| skewness bias | measue of the degree to which data "leans" towards one side |
| measurement bias | prejudice in data that represents when the sample is not representative of the actual sample population |
| to produce unbiased results | the sample tested must be sufficiently random |
| information bias | prejudice in data tha results when either the respondent or interviewer has an agenda, is not impartial or truly honest |
| response bias | respondants say what they believe the questioner wants to hear |
| response bias | can occur as a result of the way a question is worded |
| conscious bias | occurs when the surveyor is actively seeking a certain reponse |
| consequences of sub-par data quality | wasted money, bad or delayed decisions, mistrust |
| wasted money | 16%-18% of business budgets are eaten up by poor data quality |
| bad or delayed decision | if you suspect you're dealing with unreliable or incomplete data, you might delay making a decision. If you don't suspsect your decision may well be wrong |
| mistrust | poor quality data often breeds this among internal departments or exernally with customers thus leading to lost production and/or sales |
| qualitative data | data not characterized by numbers but rather are texutal, visual or oral |
| qualitative data | focused on stories, visual portrayals, meaningful characterizations, interpretations & other expressive descriptions |
| quantitive data | quantity or measurement |
| quantitive data | represents phenomena by assigning numbers in an ordered and meaningful way |
| qualitative research | primarly exploratory reasearch |
| qualitative research | used to gain an undrestanding of underlying reasons, opinions & motiviations |
| qualitative research | provides insights into the problem or helps to develop ideas or hypothesis for potential quantitiave research |
| observational study | used beuase it is impractical or impossible to control the conditions of the sutdy |
| observational study | response variables can be observed within their natural environment, giving the sense that they haven't been artificially constrained |
| observational study | generally considered weaker in terms of statisitical inference |
| quantitive research | used to quantify the problem by generating numerical data that can be transformed into useable stats |
| elements of experimental study | experimental units, treatments & responses |
| experimental units | the subjects or objects under obseration |
| treatments | the procedures applied to each subject |
| responses | the effects of the experimental treatments |
| common purpose: qualitative research | discover ideas, used in exploratory research w/general research objects |
| common purpose: quantitive research | test hypotheses or specific research questions |
| approach: qualitative research | observe & interpret |
| approach: quantitative reaearch | measure & test |
| data collection approach: qualitative | unstructured, free-form |
| data collection approach: quantitative | structured response, categories provided |
| researcher independence: qualitative | researcher is intimately involved, results are subjective |
| researcher independence: quantitive | researcher is uninvolved observer, results are objective |
| samples: qualitiatve research | small samples - often in natural settings |
| samples: quantitive research | large samples ot produce generalizable results (results that apply to other situations) |
| most often used: qualititave reasearch | exploratory research designs |
| most often used: quantitative reasearch | descriptive & causal research designs |
| big data has very little use | if we don't use tools to make it manable and applicable in some way to help make business decisions |
| big data | usually requires the use of a computer to manage |
| interval data | can not be mutliplied or divided |
| interval data | examples include time, date, temps |
| ratio data | all analysis techniques can be used with this data type |
| example of descriptive type of analysis | median household income in new york is $58,000 |
| example of predictive type of analysis | alternating 2 EMT teams between 5p & 10p will improve emergency response time |
| example of prescriptive type of analysis | employee turn over for an organization will be 5% |
| example of nominal data | types of cars produced in a factory |
| example of ratio data | distance between LA & san francisco because it can be less than whole miles |
| example of ordinal data | age of survey participants because it's putting it on a scale |
| example of ratio data | total sales for the day in dollars because it can be less than whole dollars |
| systematic error | scale that doesn’t return to 0 after each weigh |
| use of data in decision making | can lead to more effective and accurate decisions/recommendations |
| use of data in decision making | proves if samples are statisitically the same or different |
| samples must be drawn randomly | to increase likelihood of sample representing population of interest |
| example of nonrandom sample | asking friends their opinion of a new product |
| situation which may create a response bias | a survey of employees conducted by their supervisor on job satisfaction |
| conscious bias example | a marking manager conducting a focus group on levi's mens suits asking "Wouldn't you like to see a levi's tab on your business suit?" |
| missing data or refusals | everyone is not represented & skews average satisfaction scores |
| small sample sizes | assumptions made might not be realistic |
| wrong tool | leads to poorer results. |
| lack of blinding | knowledge of which group is receiving which treatment |
| associating causality example | if an increase in stork's nest occurs in a village where there is an inrease in childbirth |
| example of when to use a control chart | manufacturing faric to see if the color of each bolt is consistent and in control |
| example of when to use trend analysis | marketing research to see if TV advertising is effective over timer |
| example of when to use regression | finding cause & effect of dietary changes on health |
| example of when to use linear programming | calculating the optimal path for a plane to fly |
| example of when to use ANOVA (3 or more comparison groups) | comparing outcomes of students' ACT scores form each state |
| example of when to use t-test | comparing airline satisfaction between business & vacation travellers |
| example of when to use chi square (categorical data) | determining if women statistically prefer watching crime series dramas more than men |
| example of when to use crossover analysis | deciding if automation is less expensive in the long-term over manual processes |
| select variables that are | observeable & measurable |
| control other independent variables | such as when testing out a new pizza variety, if some respondants are hungry, they will skew results |
| identify all | assumptions made in the study |
| check to see if any | biases existed in the research |
| look for other research in this area | to build on their concepts & to see if results between studies are consistent or not (and understand why or why not) |
| cite and utilize | credible sources |
| mean | mathmetical average |
| median | value in the center of the data range (freeway median) |
| mode | most frequent occurance |
| when mode, median & mean are the same number | a true normal distribution has occurred |
| mean | most sophisticated of the 3 measures of central tendance |
| measures of central tendency | mean, mode & median |
| unlinke median & mode, mean is | influenced by the size of the values in the dataset |
| extreme values or outliers have a greater influence on | mean than they do on median or mode |
| mean can only be calculated for | interval or ratio data |
| x-bar | symbol for the sample mean |
| symbole for population mean is | a weird looking u |
| median | the point at which an equal number of scores fall above & below |
| median | the half-way point of the data |
| median | an equal number of values in a distribution are greater than the median & less than the median |
| median | calcuated by taking the average of numbers if you have 2 data points |
| median | if you have an even number of values, this is the halfway point between the 2 middle data points |
| mode | least sophicsticated of the 3 central tendancy measures |
| mode | single score or value occuring most often |
| mode | there can be more than one of these |
| mode | can't be used in inferential statistics |
| mode | is not a mathmatical computation |
| excel median | MEDIAN(b2:b16) |
| excel mode | MODE(B2:B16) |
| excel mean | AVERAGE (b2:b16) |
| variance formula | uses weird 0 squared |
| standard deviation formula | uses regular 0 and square root of formula |
| interquartile range | measures the difference between the 3rd & 1st quartile |
| to determine the quartiles | order the data from lowest to highest then separate it into 4 equal groups |
| excel quartiles | sort quartiles form lowest to highest then use quartile command (ex: for quartile 1 QUARTILE(a:2:a13,1) |
| box plot | also known as a box and whiskers or hinge plot |
| box plot | useful when showing non-parametric data as it takes into account median and percentiles rathe than averages |
| line graph | plots the relationship between 2 or more variables by using connected data points |
| line graph | very usueful when there is time series data to be summarized |
| line graph | appropriate where the data is continuous |
| vertical bars in a histogram | show the counts or numbers in each range |
| histogram | graph that displays continous data |
| a comparison of the ranges on a histogram | helps the audience understand the information presented |
| bar chart | summarizes categorical data |
| bar chart | data uses a number of bars with the same with, each representing a particular category |
| histogram | measures how continuous data is distributed over various ranges |
| bar chart | measures data that is distributed over groups or categories. |
| histogram | usueful in identifying a distribution of data |
| bar charts | more effective for understanding, weighting or quantiy of items by category |
| the wider the histograph distribution | the greater the variance of the data |
| historgram distributions | can have the same mean but different variances |
| histogram | the shape of the graph illustrates how the data is distributed |
| scatter diagram: if a correlation doesn't exists | the data points on the diagram line up along a curve or straight line across the chart |
| scatter diagram: if a correlation exist | the data does not fall along a curve or a line |
| higher R score | the more highly correlated the data |
| R value close to 1 | data is highly correlated |
| R value close to 0 | data is not correlated |
| scatter diagrams or plots | tools that look for relationships between 2 variables |
| scatter diagrams or plots | not statistically significant such as usage of regression |
| T-tests: paired sample | same unit/person before & after treatment, looking for a significant difference |
| example paired sample t-test | blood pressure of 1 subject is taken before and after 30 minutes of harp playing to see if there is a difference |
| 1 sample t-test | comparing sample mean to a target |
| example of a 1 sample t-test | fire chief compares time it takes firement to get into their truck after bell rings against a 2 min standard. |
| independent sample t-test | AKA 2 sample t-test or student's t-test, comparing mutliple sample groups |
| example of independent sample t-test | department head of PT/OT compares hours of appointments per FTE between groups to see if there is a difference in productivity |
| independent sample t-test | equal number of participants in each sample is NOT required |
| paired sample t-test | equal number of participants in each sample is IS required |
| independent sample t-test | as a rule of thumb 15 or more participants should be in each group |
| chi square | applied when you have 1 or more categorical variables from a single pouplation |
| chi square | used to determine whether there is a significant association between the 2 variables. |
| use chi square | in an election survey, voters might be classified both by gender and voting party |
| use chi square | to determine whether gender to related to voting preference |
| chi square | is basically just a table |
| one way chi square | only 1 dimension/variable is referenced |
| null hypothesis | the number noted should be approximately the same across all time periods |
| standard deviation | is the square root of the variance |
| mean | can be effected by outliers |
| interquartile range | not affected by outliers |
| 1st & 4th quartlies | where ouliers would be if any |
| box plot | line is outer or 1st & 4th quartiles |
| box plot | boxes are inner quartiles |
| scatter plot: x axis (horizontal - sex) | independent variable |
| scatter plot: y axis (vertical) | dependent variable or the result or prediction of how it will be affected by the other axis |
| scatter plot | best way to confirm relationship is to run a regression on the points and determine the p value |
| p value less than .05 | there IS a significate relationship between factors |
| p value more than .05 | there is NOT a significate relationship between factors |
| paired sample t-test | looks at the same people before AND after a treatement, looking for a treatment effect |
| linear regression | determining if there is a statistically significant relationship between a target (dependent) variable & 1 or more predictor (independent) variables |
| target variable | dependent variable |
| predictor variable | independent variable |
| linear trend analysis | determing if there is a trend over time |
| multiple regression | determing if mutliple variables have a significatn relationship with dependent variable |
| probability | the chance of an event occuring at some point in the future |
| probability | involves chance & outcomes and can be based on independence or associations |
| probability | is between 0% & 100%, between 0 and 1 |
| probability | helps to determine odds of future outcomes occuring |
| probability | helps in the assessment of what decisions to make based on likelihood of future trends |
| probability | whas it the risk of doing nothing? What is the chance of failure? How likely are we in the achievement of our market share goals? |
| probability | Can assist in risk maangement. What is the chance of a highly adverse outcome occuring? |
| probability | common examples include: a flip of a coin, rolling a die or drawing a certain card from a deck. |
| probability | numer of possiblities that meet conditions/ number of equally likely possiblities |
| probability combination principle | total outcome for several events = individual's # times # of possible outcomes for each additional event EX. 3 dice = 6 *6*6 or 216 possible outcomes |
| probability sample with replacement | an item is removed from the group and then added back. Ex: trying on clohting then putting it back on the rack |
| probability sample without replacment | an item is removed but not put back. Ex: lottery ball kept out when drawn |
| probability with overlap | make the interlocking circle diagram - add items in circle and subtract # in overlap |
| measures of dispersion | range, variance & standard deviation |
| variance | measure of how spread out data are about the mean |
| standard deviation | measure of how far, on average, the data points are from the mean. |
| on mean/standard deviation chart | each verticle line represents an incremental standard deviation from the mean |
| z-score | score relating to confidence, is equivelent to one standard deviation from the mean. |
| z = plus or minus 0.674 | 50% confidence |
| z = plus or minus 1.645 | 90% confidence |
| z = plus or minus 1.960 | 95% confidence |
| z = pluse or mins 2.576 | 99% confidence |
| normal distribution graph | shows mean & standard deviations (ghost with lines through it) |
| normal distribution graph | mean of the distribution determins the location of the center of the graph |
| large standard deviation & variant | short and fat ghost |
| small standard deviation & variant | tall and skinny ghost |
| normal distribution graph | area under the curve represents the probability of events happening |
| normal distribution graph | total area under the normal curve, as with any distribution is 1 |
| cumulative probability | sum of probabilities |
| in relation to normal distribution, a cumulative probability refers to | the probability that a randomly selected score will be less than or equal to a specified value |
| z-score | statistical measurement of a score's relationship so the mean |
| z-score of 0 | the score is the same as the mean |
| z-score | can be positive or negative, indicating wither it is above or below the mean & by how many standard deviations |
| linear trend analysis | independent variable or x is always times |
| linear regression analysis | x value is never timed, is a different variable |
| alternate hypothesis | there IS a significate relationship between factors |
| null hypothesis | there is NOT a significate relationship between factors |
| if P value for the independent value is less than .05 | you reject the null hypothesis & accept alternate hypothesis |
| addition rule of probability | question asks for probability of 1 OR another event occuring. Add the chances together |
| mutually exclusive | events that CAN NOT occur at the same time. Ex. Rolling a 1 AND 6 on 1 die |
| addition rule of probability | if more than 1 chance, add all of the chances together |
| Bayes Theorem | conditional probability |
| Bayes Theorem | "given that" usually used in the problem |
| Bayes Theorem: P(spots) | P(Dog) * P(Spots/Dog) + P(Cat) * P(spots/cat) = |
| normal distribution graph | 2/3 or 62% of the points are 1Z or 1 standard deviation below the mean & 1 above the mean |
| normal distribution graph | the 1st deviations are the 2 on either side of the mean |
| normal distribution graph - 1Z | one place to the left of the mean |
| normal distribution graph + 1Z | one place to the right of the mean |
| 250 bank customers with 8.2 min avg wait time, 1.2 standard deviation, 94% prob customer waiting | want to be 2 standard deviations from the mean. Multiply standard deviations by number given. 1.2 *2. Add to mean for positive and subtract from mean for neg |
| tools used to analyze business | linear programming, breakeven & crossover analysis, normal distribution & ANOVA, forecasting, cluster analysis, decision analysis, chi-square |
| objective of all linear programming problems | the maximization or minimization of some quantity |
| all linar programming problems | have constraints that limit the degree to which the objective can be pursued |
| a feasible solution | satisfies all the problem's constraints |
| an optimal solution | a feasbile solution that results in the largest possible objectie fucntion value when maximizing (or smallest when minimizing) |
| a graphical solution method | can be used to solve a linear program with two variables |
| linear programming problem | both the objective function & the contraints are linear |
| linear functions | functions in which each variable appears in a separate term raised to the first power and is multiplied by a constant (which could be 0) |
| Linear constraints | are linear functions that are restricted to be "less than or equal to", "equal to", or "greater than or equal to" a constant. |
| linear programming problem | feasible region is the region below both lines, farthest to the bottom left |
| linear programing extreme points | points located on the line that surrounds the feasible region,best on corners |
| linear programming analysis | helps with mutliple variables (items) with mutliple constraints |
| linear programming | we can consider far more than 2 items |
| linear programming | we use software that does the math |
| linear programming | very effective way to find the best solution for mutliple products w/mutliple limitations to reach a "best" goal |
| breakeven & crossover analysis | analyze trends to determine when 2 characteristics are equal |
| breakeven & crossover analysis | consider when profits will be at a break even point? At what volume will revenues & costs be equal? |
| breakeven & crossover analysis | At what volumes are 2 approaches equal? At what volume are total costs equal & when is one better than the other? |
| breakeven graph | where revenue & volume cross each other is your breakeven |
| breakeven & crossover analysis | used to analyze trends to determine when 2 characteristics are equal |
| breakeven | usually used with profits & therefore includes revenues & costs |
| crossover | usually looks at various options to determine when a particular option is most attractive & when another option is better |
| normal distribution graph | the higher the curve at any point, the more likely that event will happen |
| mean (average) | most liekly outcome of the "bell curve" |
| standard deviation | the distance around the mean where 2/3 of events happen |
| standard eviation | determines (describes) the height & width of the normal deviation graph |
| large standard deviation & variant | process is highly variable, wide range of things happen on a routine basis |
| small standard deviation & variant | process is highly uniform, almost always happens |
| standard normal distribution | has a mean of 0 & a standard deviation of 1 |
| standard normal distribution | all results are above or below the mean or ZERO |
| average temp in oct is 65, standard deviation of 10 | mean = 65, -1 standard devation = 55 & 1 standard deviation = 75 |
| area under the curve on normal distribution graph | represents the probability of events happening |
| area under the curve on normal distribution graph | total area under the nomal curve, as with any distribution, is 1 or 100% |
| the emperical rule: normal curve probabilities | approximately 2/3 of data points within a dataset will be within 1 standard deiation of the mean |
| approximately 95% of all data poitns | will be within 2 standard deviations of the mean |
| Almost all (99%) o the data points | will be within 3 standard deviations of the mean |
| normal distribution graph | or normal curve |
| normal distribution graph | mean nearly equal to median |
| multip modal distribution | mutliple hills |
| bimodal distribution | 2 hills |
| normal distrubtion | 1 hill |
| pareto | one smooth skate ramp and others with sharp points |
| ANOVA | statistical analysis to determine whether separate data sets are different form each other or whether they are really too similar to be different |
| ANOVA null hypothesis | AKA straw model or basic assumption |
| ANOVA null hypothesis | all mean populations are equal |
| ANOVA alternate hypothesis | not all of the population means are equal but at aleast one are significantly different |
| ANOVA F-value | test statistic ussed to set the region in which we can reject the null hypothesis & show that the alternative hypothesis is accepted |
| if the ANOVA null hypothesis is false | F-value is likely to lie in the region |
| if the ANOVA null hypothesis is false | at least one of the population means is significantly different from others, subject to the stated level of significance |
| linear programming graph best answers | one of the corners along the feasible area parimeter |
| to find the best of the best answers from linear programming graph | plug coordinates at corners of feasible into problem & select one that gives me highest number/profit |
| crossover analysis | only considers cost not revenue |
| breakeven analysis | considers both cost & revenue |
| crossover graph: crossover point | where 2 criteria are equal or cross |
| six sigma | don't want to see any events occuring more than + or - 6 standard deviations due to extremely rare or unlikelines of occurance |
| standard curve | usually no more than + or - 3 standard deviations |
| culumlative probability | on graph everthing blue if one line or point is listed or the entire area meeting that criteria |
| 1/6th | is to the left of 1 standard deviation & to the right of 1 standard deviation or about 16.7% |
| chance of six sigma occuring | 3.4 out of 1mil chances |
| six sigma | graph goes up to 6 standard deviations |
| in quality if something is 3 or more standard deviations | indicates there is a problem in the process that likely needs to be address |
| F critical | usually between 2-4 |
| if f value is higher than f critical value | reject null, accept alternative |
| forecasting | attempting to predict the future and/or events that will happen in a different situation to make the best possible decistion |
| in order to forecast | determine whether there are patters in past events & determine which info is most likely to lead to accurae forecast |
| forecasting techniques | regression analysis, time series analysis, cluster analysis, decision analysis |
| regression analysis technique | find a best-fit line thorugh data points that most closely matches points |
| regression analysis benefits | allows sophisticated analysis of cost behavior sales forecasts |
| regression analysis benefits | provides objective benchmarks for evaluation of realiability of estimates |
| regression analysis disadvantages | requires 15 or more data points for accuracy |
| regression analysis disadvantages | can be influeced by outliers |
| regression analysis disadvantages | requires informed analysis |
| Time series analysis technique | Use past results to predict events that will occur in the future |
| Time series analysis benefits | Aids decision making by finding patterns in data, such as sales trends |
| Time series analysis benefits | Allows performance and productivity evaluation |
| Time series analysis shortcomings | Assumes past data patterns will repeat in future, which may not be true |
| Time series analysis shortcomings | Key variables may not be captured |
| Cluster analysis technique | Plot a series of data points and look for trends or patterns that increase our understanding |
| Cluster analysis benefits | Sorts individual data points into different groups |
| Cluster analysis benefits | - Helps determine target markets |
| Cluster analysis benefits | Identifies successful and unsuccessful habits and systems |
| Cluster analysis disadvantages | Long and expensive process |
| Cluster analysis disadvantages | There are hundreds of potential approaches to take, each specific to a certain situation |
| Decision analysis technique | Organized analysis of a series of decisions, events, and the value of those outcomes to determine the decision most likely to give us the best outcome. |
| Decision analysis technique | A Decision Tree is one example. |
| Decision analysis benefits | Determines the decision with the greatest value |
| Decision analysis benefits | Produces a value under certainty, uncertainty, and risk |
| Decision analysis disadvantages | Quality of decision is limited to the amount of data available |
| Decision analysis disadvantages | Does not emphasize the risk of the worst case scenario |
| Regression Analysis | Fitting a trend line to historical data points to project into the medium to long-range |
| Regression Analysis: Linear trends | can be found using the least squares technique |
| regression analysis: least squares method | minimizes the sm of the squared errors (deviations) |
| least squares method map | straight trend line with deviations hanging onto it (above and below) |
| multiple regression | when regression has more than 1 independent variable |
| multiple regression | hard to graph more than 3 dimensions |
| logistic regression used for | categorical variables (on/off, brand A/Brand B) & non-linear regression line (ie curved) |
| autocorrection | occurs when a given daata point on a time series analysis is affected by a previous data point for that time serires |
| in ordinary regression analysis | we assume that errors are independent from one another |
| Autoregressive Error Correction | produces a superior regression analysis compared to ordinary regression analysis because it takes autocorrelation into account. |
| autocorrelation | If the previous day was sunny and hot, it is not very likely it will snow that day. It is more likely if there was snow on the previous day. |
| Homoscedasticity: | The Variability of the data is similar for all values of the variables |
| Heteroscedasticity: | The Variability of the data changes as we move through different values of the variables. |
| Forecasts | are seldom perfect |
| Most techniques | assume an underlying stability in the system |
| Product family and aggregated forecasts | are more accurate than individual product forecasts |
| time-series models | 1. Moving average 2. Weighted Moving averages 3. Exponential smoothing 4. Trend projection |
| associative model | Linear regression |
| Time Series Forecasting | Set of evenly spaced numerical data |
| Time Series Forecasting | Obtained by observing response variable at regular time periods |
| Time Series Forecasting | Forecast based only on past values, no other variables important |
| Time Series Forecasting | Assumes that factors influencing past and present will continue influence in future |
| Time Series Components | trend, cyclical, seasonal, random |
| time seris gragh - average demand | average demand horizontal line |
| time series graph - trend compoenet | upward or downward straight line |
| time series graph - actual demand | wildly swinging up and down line |
| Cluster analysis | also known as segmentation |
| Cluster analysis | the process of arranging terms or values based on different variables into "natural" groups |
| Most often with cluster analysis | these terms or values are survey responses from people. |
| cluster analysis | There are hundreds of approaches& it is used in many different fields to have a better understanding of an industry's environment. |
| Cluster Analysis | often a trial-and-error approach to find natural groupings |
| Decision analysis | can be used to develop an optimal strategy when a decision maker is faced with several decision alternatives and an uncertain or risk-filled pattern of future events. |
| Even when a careful decision analysis has been conducted | the uncertain future events make the final consequence uncertain. |
| The risk associated with any decision alternative is | a direct result of the uncertainty associated with the final consequence |
| A good decision analysis includes | risk analysis that provides probability information about the favorable as well as the unfavorable consequences that may occur. |
| A decision problem is characterized by | decision alternatives, states of nature, and resulting payoffs. |
| decision alternatives | are the different possible strategies the decision maker can employ. |
| states of nature | future events, not under the control of the decision maker, which may occur |
| states of nature | should be defined so that they are mutually exclusive and collectively exhaustive. |
| Expected Value Approach | If probabilistic information regarding the states of nature is available, one may use this |
| Expected Value Approach | the expected return for each decision is calculated by summing the products of the payoff under each state of nature and the probability of the respective state of nature occurring. |
| Expected Value Approach | The decision yielding the best expected return is chosen. |
| PDC decision tree: expected values | Choose the decision alternative with the largest EV. |
| Chi-Square Test | Used with frequency of events |
| Chi-Square Test | The number of events, not percentages |
| Chi-Square Test | Event counts for each cell must be greater than 5 |
| Compare our statistical test results with | target number in a Chi-Square table |
| Chi-Square Test - Null Hypothesis | often not very exciting |
| Chi-Square Test - Null Hypothesis | no relationship between event variables |
| Chi-Square Test - Null Hypothesis | relationship follows known theory or fact |
| Chi-Square Test - Null Hypothesis | Choose a statistical significance (p-value) |
| Chi-Square Test: a p-value of 5% | means that the data we found would support the Null Hypothesis only 5% or less of the time, and that we can be 95% confident that we can reject the Null Hypothesis |
| t-test & ANOVA | the same process pretty much |
| if p is low | null must go |
| t-test | 2 samples |
| ANOVA | more than 2 samples |
| time series | use when time periods are involved |
| linear regression | same as least squares |
| logistics regression better choice | when dealing with curved line or categories |
| exponential smoothing | fancy way to do weighted moving averages but it's commonly used |
| time series models | set of evenly spaced numerical data tracked on a consistent time frame ie quarterly whose forecast is based solely on past data |
| time series models trend adjustment | an adjustment to make sure that if sales are growing we are not underestimating by looking at past sales that are lower |
| time series models cyclical adjustment | adjustment made based on business cycles such as recessions or boom times, related to national economic situation |
| time series models seasonal adjustment | calendar based adjustment EX. If most of our sales are in December we will adjust December's forecasts |
| time series models random adjustment | adjustment made based on random events such as a competitor going out of business or having a black Friday sale |
| cluster analysis | often used in biology and many other sciences |
| cluster analysis | usually based on survey data, used a lot by marketing |
| cluster analysis | cares most about what characteristics belong to each group and then decides how to interact with those groups |
| decision tree: weighted average states of nature | multiply each state's percentage by that state's value then add the state's totals together for each component. Choose most money |
| Quality Management Principles | customer focus, leadership, continual improvement, system approach to management, mutually beneficial supplier relationship, process approach, factual approach to decision making, people involvement |
| Quality Management Principle: Customer focus | Organizations depend on their customers and therefore should understand current and future customer needs, should meet customer requirements and strive to exceed customer expectations. |
| QM customer focus benefit | Increased revenue and market share obtained through flexible and fast responses to market opportunities |
| QM customer focus benefit | Increased effectiveness in the use of the organization’s resources to enhance customer satisfaction |
| QM customer focus benefit | Improved customer loyalty leading to repeat business. |
| Quality Management Principle: Leadership | Leaders establish unity of purpose and direction of the organization. They should create and maintain the internal environment in which people can become fully involved in achieving the organization’s objectives. |
| QM leadership benefit | People will understand and be motivated towards the organization’s goals and objectives |
| QM leadership benefit | Activities are evaluated, aligned and implemented in a unified way |
| QM leadership benefit | Miscommunication between levels of an organization will be minimized. |
| Quality Management Principle: continual improvement | Continual improvement of the organization’s overall performance should be a permanent objective of the organization. |
| QM continual improvement benefit | Performance advantage through improved organizational capabilities |
| QM continual improvement benefit | Alignment of improvement activities at all levels to an organization’s strategic intent |
| QM continual improvement benefit | Flexibility to react quickly to opportunities. |
| Quality Management Principle: system approach to mgmt | Identifying, understanding and managing interrelated processes as a system contributes to the organization’s effectiveness and efficiency in achieving its objectives. |
| QM system approach to mgmt benefit | Integration and alignment of the processes that will best achieve the desired results |
| QM system approach to mgmt benefit | Ability to focus effort on the key processes |
| QM system approach to mgmt benefit | Providing confidence to interested parties as to the consistency, effectiveness and efficiency of the organization |
| Quality Management Principle: mutually beneficial supplier relationship | An organization and its suppliers are interdependent and a mutually beneficial relationship enhances the ability of both to create value |
| QM mutually beneficial supplier relationship benefit | Increased ability to create value for both parties |
| QM mutually beneficial supplier relationship benefit | Flexibility and speed of joint responses to changing market or customer needs and expectations |
| QM mutually beneficial supplier relationship benefit | Optimization of costs and resources. |
| Quality Management Principle: process approach | A desired result is achieved more efficiently when activities and related resources are managed as a process. |
| QM process approach benefit | Lower costs and shorter cycle times through effective use of resources |
| QM process approach benefit | Improved, consistent and predictable results |
| QM process approach benefit | Focused and prioritized improvement opportunities. |
| Quality Management Principle: factual approach to decision making | Effective decisions are based on the analysis of data and information |
| QM factual approach to decision making benefit | Informed decisions |
| QM factual approach to decision making benefit | An increased ability to demonstrate the effectiveness of past decisions through reference to factual records |
| QM factual approach to decision making benefit | Increased ability to review, challenge and change opinions and decisions. |
| Quality Management Principle: people involvement | People at all levels are the essence of an organization and their full involvement enables their abilities to be used for the organization’s benefit. |
| QM people involvement benefit | Motivated, committed and involved people within the organization |
| QM people involvement benefit | Innovation and creativity in furthering the organization’s objectives |
| QM people involvement benefit | People being accountable for their own performance |
| QM people involvement benefit | People eager to participate in and contribute to continual improvement. |
| Shewhart's PDCA Model | Plan-Do-Check-Act |
| Shewhart's PDCA Model: Plan | Identify the pattern and make a plan |
| Shewhart's PDCA Model: Do | Test the plan |
| Shewhart's PDCA Model: Check | Is the plan working? |
| Shewhart's PDCA Model: Act | Implement the plan document |
| Supplier-Input-Process-Output-Customer (SIPOC) | a high-level view of a process |
| SIPOC: S supplier | person/org that provices input to a process |
| SIPOC: I input | resource that is added to a process by a supplier |
| SIPOC: P process | series of steps where an input converts to an output |
| SIPOC: O output | resource that is the result of a process |
| SIPOC: C customer | person/org that receives products or services |
| quality assurance | An overall management plan to guarantee the integrity of data (The “system”) |
| quality control | A series of analytical measurements used to assess the quality of the analytical data (The “tools”) |
| quality assurance | prevention |
| quality control | detection |
| quality control: focus | Uncover defects so they can be fixed |
| quality assurance: focus | Prevent defects from occurring |
| quality control: purpose | Assess performance and recommend corrective action |
| quality assurance: purpose | Assess capability and recommend preventive action |
| quality control: level | Basic—recognize problems so they can be fixed |
| quality assurance: level | Advanced—understand the intricacies of the system and predict outcomes |
| quality control: major activities | Inspection and repair |
| quality assurance: major activities | Design and Training |
| quality control: change response | Reactive—take action once the problem has occurred |
| quality assurance: change response | Proactive—take action before the problem can occur |
| quality control | C - average, basic |
| quality assurance | A - superior, proactive |
| Seven Tools of TQM Total Quality Management) | Check sheets, Scatter diagrams, Cause-and-effect diagrams, Pareto charts, Flowcharts, Histogram, Statistical process control chart |
| Check Sheet: | An organized method of recording data (looks like a table with tick marks) |
| Scatter Diagram | A graph of the value of one variable vs. another variable (unconnected dots on graph) |
| Cause-and-Effect Diagram | A tool that identifies process elements (causes) that might effect an outcome |
| Cause-and-Effect Diagram | looks like a decision tree with the branches on the left (cause) & box (effect) on right |
| Flowchart (Process Diagram): | A chart that describes the steps in a process |
| Histogram | A distribution showing the frequency of occurrences of a variable |
| Histogram | looks like a bar chart in bell curve shape |
| Pareto Chart | A graph to identify and plot problems or defects in descending order of frequency |
| Pareto Chart | graph starts high and then tails off, can be lines or bars with a ball chain dotted line over them |
| Statistical Process Control Chart | A chart with time on the horizontal axis to plot values of a statistic |
| Statistical Process Control Chart | chart with a solid target value line and dotted upper & lower control limit lines |
| Statistical Process Control (SPC) | Uses statistics and control charts to tell when to take corrective action |
| Statistical Process Control (SPC) | Drives process improvement |
| Statistical Process Control (SPC): 4 key steps | Measure the process |
| Statistical Process Control (SPC): 4 key steps | When a change is indicated, find the assignable cause |
| Statistical Process Control (SPC): 4 key steps | Eliminate or incorporate the cause |
| Statistical Process Control (SPC): 4 key steps | Restart the revised process |
| Statistical Process Control (SPC) | Variability is inherent in every process: Natural or common causes AND/OR Special or assignable causes |
| Statistical Process Control (SPC) | Provides a statistical signal when assignable causes are present |
| Statistical Process Control (SPC) | Detect and eliminate assignable causes of variation |
| An SPC Chart Types of Data: Variables | For variables that have continuous dimensions such as Weight, speed, length, etc. |
| x-charts | are to control the central tendency of the process |
| R-charts | are to control the dispersion of the process |
| An SPC Chart Types of Data: Attributes | Defect-related characteristics |
| An SPC Chart Types of Data: Attributes | Classify products as either good or bad or count defects |
| An SPC Chart Types of Data: Attributes | Categorical or discrete random variables |
| SPC chart: anything outside of the upper & lower limits | is out of control |
| SPC chart: anything outside of the upper & lower limits | variation due to assignable causes |
| SPC chart: anything inside the upper & lower limits | variation due to natural causes |
| Control Charts for Attributes | For variables that are categorical such as: Good/bad, yes/no, acceptable/unacceptable |
| Control Charts for Attributes | Measurement is typically counting defectives |
| Control Charts for Attributes | Charts may measure: Percent defective (p-chart) or Number of defects (c-chart) |
| Control Charts for Attributes: P-chart | measures percent defective (p percent) |
| Control Charts for Attributes: C-chart | measures number of defects (c count) |
| The Seven New Tools for Improvement | affinity diagram, tree diagrams, interelationship digraph, process decision program chart, activity network diagram, matrix diagram, prioritization grid |
| New tools to analyze non-numerical data: affinity diagram | A tool that groups items based on relationships which are then analyzed |
| New tools to analyze non-numerical data: affinity diagram | Brainstorming tool that organizes large amounts of disorganized data and information into groupings based on natural relationships. |
| New tools to analyze non-numerical data: affinity diagram | used when: You are confronted with many facts or ideas in apparent chaos. |
| New tools to analyze non-numerical data: affinity diagram | used when: Issues seem too large and complex to grasp. |
| New tools to analyze non-numerical data: affinity diagram | columns with a category heading with "stickies" under each heading |
| New tools to analyze non-numerical data: Interrelationship digraph | info in boxes stuck in a perimeter with arows drawn between boxes in different colors denoting a relationships between boxes |
| New tools to analyze non-numerical data: Interrelationship digraph | This tool displays all the interrelated cause-and-effect relationships and factors involved in a complex problem and describes desired outcomes. |
| New tools to analyze non-numerical data: Interrelationship digraph | The process of creating an interrelationship digraph helps analyze the natural links between different aspects of a complex situation. |
| New tools to analyze non-numerical data: Interrelationship digraph | The process of creating an interrelationship digraph helps analyze the natural links between different aspects of a complex situation. |
| New tools to analyze non-numerical data: Tree diagram | basically a flow chart with double slanted lines coming out of subordinate boxes |
| New tools to analyze non-numerical data: Tree diagram | A hierarchical tool that breaks a topic down into its components |
| New tools to analyze non-numerical data: Tree diagram | Breaks down broad categories into finer and finer levels of detail. |
| New tools to analyze non-numerical data: Tree diagram | Developing a tree diagram directs concentration from generalities to specifics. |
| New tools to analyze non-numerical data: Prioritization matrix | grid table |
| New tools to analyze non-numerical data: Prioritization matrix | A table or chart that helps a team prioritize multiple options, based on how well these options satisfy preselected criteria |
| New tools to analyze non-numerical data: Prioritization matrix | Prioritizes items in terms of weighted criteria. |
| New tools to analyze non-numerical data: Prioritization matrix | It uses a combination of tree and matrix diagramming techniques to do a pair-wise evaluation of items and to narrow down options to the most desired or most effective. |
| Popular applications for the Prioritization Matrix | include return on investment(ROI) or Cost - Benefit Analysis (investment vs. return), time management matrix (urgency vs. importance), etc |
| New tools to analyze non-numerical data: Matrix diagram | grid table that uses symbols |
| New tools to analyze non-numerical data: Matrix diagram | A table or chart that shows the strength of the relationships between items or sets of items |
| New tools to analyze non-numerical data: Matrix diagram | At each intersection a relationship is either absent or present. |
| New tools to analyze non-numerical data: Matrix diagram | It then gives information about the relationship, such as its strength, the roles played by various individuals or measurements. |
| New tools to analyze non-numerical data: Process decision program chart | Like a tree diagram, the PDPC shows a hierarchy of events or ideas but the intent of the PDPC is more defined |
| New tools to analyze non-numerical data: Process decision program chart | the lowest tier of the PDPC illustrates the corrective and preventive actions that can be taken to mitigate risks or overcome process problems. |
| New tools to analyze non-numerical data: Process decision program chart | Specifically designed to help teams mitigate risks and solve potential problems. |
| New tools to analyze non-numerical data: Process decision program chart | Different shaped boxes are used to highlight risks and identify possible countermeasures |
| PDPC | Process decision program chart |
| New tools to analyze non-numerical data: Network diagram | A decision diagram with a green ball on the left & a red ball on the right |
| New tools to analyze non-numerical data: Network diagram | A scheduling diagram that shows the relationships between project activities |
| New tools to analyze non-numerical data: Network diagram | It is used when subtasks occur in parallel. |
| New tools to analyze non-numerical data: Network diagram | The diagram helps in determining the critical path (longest sequence of tasks) |
| critial path | longest sequence of tasks |
| TQM | Encompasses entire organization, from supplier to customer |
| TQM | Stresses a commitment by management to have a continuing, companywide drive toward excellence in all aspects of products and services that are important to the customer |
| TQM encompasses | Continuous improvement, Statistical Quality Control, Employee empowerment, Benchmarking, Just-in-time (JIT), Knowledge of TQM tools |
| Lean Operations | are externally focused on the customer |
| Lean Operations | Emphasis on understanding the customer and what the customer wants |
| Lean Operations | Optimizes the entire process from the customer’s perspective |
| Lean classifies every activity that we do into three types | value add, non value-add but essential & waste |
| Lean Ops: Value Add | activities that a customer would be willing to pay for which help create the final form or function of the finished article |
| Lean Ops: Non Value Add but Essential | things that need to be done, but that don’t bring any value to the finished article (e.g. waiting for a document to print, the time it takes for paint to dry etc.) |
| Lean Ops: Waste | actions that bring no value to the article and are therefore unnecessary |
| Just-In-Time (JIT) | Powerful strategy for improving operations |
| Just-In-Time (JIT) | Materials arrive where they are needed when they are needed |
| Just-In-Time (JIT) | Identifying problems and driving out waste reduces costs and variability and improves throughput |
| Just-In-Time (JIT) | Requires a meaningful buyer-supplier relationship |
| Six Sigma | Statistical definition of a process that is 99.9997% capable, 3.4 defects per million opportunities (DPMO) |
| Six Sigma: DPMO | defects per million opportunities |
| Six Sigma | A program designed to reduce defects, lower costs, and improve customer satisfaction |
| Six Sigma graphing | uses bell curve graph |
| Six Sigma graphing | 3 standard deviations is 2,700 defects/million |
| Six Sigma graphing | 6 standard deviations are 3.4 defects/million |
| Lean Six Sigma | It combines the streamlining and waste-elimination concepts of Lean practices with the variation- and qualitycontrol ideas of Six Sigma |
| Combining Lean's focus on enhancing customer value with Six Sigma's optimization of process work | simultaneously reduces inefficiency, accelerates production, and increases quality |
| Design for Six Sigma (DFSS) methodology | does not wait to correct inefficiencies in processes |
| Design for Six Sigma (DFSS) methodology | it incorporates Six Sigma practices into the design of the processes |
| By taking a proactive approach to quality management, DFSS | ensures that variation is minimized from the outset of a project, eliminating the need for corrective actions later in project work. |
| Reliability | Generally defined as the ability of a product to perform as expected over time |
| Reliability | Formally defined as the probability that a product, piece of equipment, or system performs its intended function for a stated period of time under specified operating conditions |
| Types of Reliability | inherent & achieved |
| Inherent reliability | predicted by product design (robust design) |
| Achieved reliability | observed during use |
| R1 = | reliability of component 1 |
| Rs = | R1 x R2 x R3 x … x Rn |
| Reliability of the process is | Rs = R1 x R2 x R3 = .90 x .80 x .99 = .713 or 71.3% |
| ISO | The International Organization for Standardization |
| The International Organization for Standardization (ISO) | established a certification program to guarantee that organizations are dedicated to quality concepts and are continually working to ensure the highest level of quality possible |
| ISO Certification | shows that an organization has a quality management system in place to monitor and control quality issues and is continuing to meet the needs of customers and stakeholders with highquality products and services. |
| International Organization for Standardization (ISO)- | mission is to promote the development of standardized products to facilitate trade and cooperation across national borders |
| International Organization for Standardization (ISO)- | Representatives from more than 146 nations |
| ISO 9000 series of standards | sets requirements for quality processes |
| Nearly half a million ISO 9000 certificates | have been awarded to companies around the world. |
| ISO 14000 series | also sets standards for operations that minimize harm to the environment. |
| ISO | Developed the quality management principles |
| 80/20 rule | 80% of issues are caused by 20% of issues |
| check sheet | counts how many times something occurs |
| cause/effect SIPOC diagram | top branches are primary issues, wings off branches are secondary issues |
| pareto chart black dots | cumulative effects of issues |
| run chart | shows data in relation to a target value |
| control chart | run chart with control limits added. |
| statistical process control chart | control chart calcuated via statistical means or processes |
| statistical process control chart | tells WHEN to take corrective action not what to do |
| statistical process control chart | early warning system |
| natural or common cause variation | in control |
| special or assignable cause variation | out of control, created by something in the process, what you're looking for on SPC chart |
| trends within control limit | can also indicate a system issue |
| x-charts | area within the control limits is shaded in |
| 7 new tools for improvement | AKA Ishikawa tools |
| order of 7 new tool use | affinity-> interrelationional -> tree-> prioritization matrix or matrix diagram -> process decision program chart -> network diagram |
| lean ops waste | from a customer perspective inventory, movement to warehousing and quality programs would be considered this |
| Just-In-Time (JIT) | internal operations perspective |
| Just-In-Time (JIT) | opposite of lean ops focus, often run in conjuction with lean ops |
| 6 standard deviations from mean | 99.997% perfection |
| R S subscripted | overall sysem reliablity |
| RBM | Results based management |
| results based management (RBM) | uses results as the central measure of performance |
| results based management (RBM) | translates goals into results, clearly defined accountability of results & requires monitoring and self-assessment |
| results based management (RBM) | Takes a life-cycle approach |
| results based management (RBM) | continuous measurement and performance evaluation; must be measureable using data |
| Steps of RBM | Resources: inputs -> activities -> results: outputs -> outcomes -> impact |
| RBM requires | partnerships and inclusiveness; shared expectations; & transparency, simplicity, and flexibility |
| Results-based managedment tools table | includes expected results, indicators, baseline data, targets, data sources, data collection methods, frequency & responsbility |
| Performance Measures | Used to measure results, effectiveness, and/or efficiency of an individual, group or the entire company |
| Performance Measures | Answer such questions as: How are we doing? What do we need to do to improve? Where is problem solving needed? |
| Performance Measures | Should be linked to a company’s goals and/or strategy e.g., goal is to gain market share by improving customer satisfaction from 70% to 80% |
| Performance Indicators | virtually anything that can be tracked and quantified, |
| Performance Indicators | examples include: Financial performance, Customer satisfaction, Quality of programs or services, Employee retention, Safety statistics, Energy consumption |
| Business Improvement Analytics: Index numbers | are a common analytic for business improvement |
| Business Improvement Analytics: Index numbers | commonly represent the change in price or quantity over time for goods and/or services EX: Consumer Price Index |
| Business Improvement Analytics: Consumer price index | “basket” of assorted consumer goods and services that are purchased by a common household |
| ConBusiness Improvement Analytics: sumer price index | watched closely because considered a main measure of inflation |
| Business Improvement Analytics: CPI | Consumer Price Index |
| Business Improvement Analytics: CPI Basket includes | communication, healthcare, education, transportation, recreation, food, housing & clothing |
| Business Improvement Analytics: Indices | are usually relative to a base period that is represented as a value of 100 |
| Business Improvement Analytics: Index graph | Average is shown in addition to actual |
| Types of Indices: | Simple Index Number, Simple Composite Index & Weighted Composite Index |
| Business Improvement Analytics: Simple Index Number | Price or quantity relative to a base period of 100 |
| Price of Big Mac in 1968=$1.60, Price of Big Mac in 2014 = $4.80; express the Simple Index Number for the 2014 price of a Big Mac, using the 1968 price as the Base Period | Answer = ($4.80 / $1.60) * 100 = 300 Note: 1968 price as a simple index = ($1.60 / $1.60) * 100 = 100 |
| Business Improvement Analytics: Simple Composite Index | an index based on a combination of items/measures without weighting any data more significantly than any other data |
| suppose a Brand Equity index is created based on 5-point ratings of Quality, Value, and Uniqueness. ; Assume the average summed ratings score is 11.5 across 300 brands | What is the Simple Composite “Brand Equity” Index for a brand with a summed rating score of 12.7? = (12.7 / 11.5) * 100 = 110.4 = 110 |
| Weighted Composite Index | similar to a Simple Composite Index except that certain variables are given more weight than others when calculating the index |
| Business Improvement Analytics: Weighted Composite Index | e.g., suppose that consumer perceptions of brand Quality and Value are more important than Uniqueness in driving brand sales |
| Business Improvement Analytics: Weighted Composite Index | Could create a “Brand Equity” Index with the following weights ; 50% weight to perceived Quality ; 30% weight to perceived Value ; 20% weight to perceived Uniqueness |
| Health Care Analytics: Comonly used metrics - Rate | frequency of an event per time period ; e.g. birth rate - # births per 1,000 people in a year |
| Health Care Analytics: Comonly used metrics - Ratio | – measure of one quantity in relation to another ; e.g., gender ratio for Alzheimer’s ~ 2: 1 females to males |
| Health Care Analytics: Comonly used metrics - Proportion | ratio of a group to the whole |
| Health Care Analytics: Comonly used metrics - Prevalence | (number cases) / (total population) ; e.g., 20% of African Americans over 75+ yrs old have Alzheimer’s |
| Health Care Analytics: Comonly used metrics - Incidence | only considers new cases |
| Health Care Analytics: Comonly used metrics - cumulative Incidence | (# new cases in particular time) / (population) |
| Health Care Analytics: incidence rate | (# new cases) / (person-time units) ; person-time units -> cumulative amount of time each person was studied |
| Education Analytics: Test Construction - Norm Referenced Tests | compare an individual to others ; e.g., standard score (Z-score) |
| Education Analytics: Test Construction - Criterion Referenced Tests | compare an individual to defined standards ; e.g., exam cut-score |
| Education Analytics: Test Construction - True Score Theory | for a test without systematic error, the observed score is the true score plus random error |
| Education Analytics: Test Construction - True Score Theory: systematic error | occurs when something unrelated to the test per se is affecting the results (e.g., jack-hammering was going on outside test center) |
| Education Analytics: Test Construction - Item Response Theory | takes into account that different questions have different levels of difficulty ; e.g., SAT and GMAT |
| Government/Public Sector Analytics | Pressures to deliver public services at lower costs ; i.e., reform spending habits and optimize the allocation of public benefits (public welfare) |
| Government/Public Sector Analytics: Cost-benefit analysis | In public sector, not as straightforward (e.g., global warming; spending with no revenue) ; Attempt to measure the benefit to the general welfare of the public |
| Government/Public Sector Analytics:; Benchmarking | e.g., anticipated cost of new transit system relative to actual cost of similar transit system of other cities |
| Government/Public Sector Analytics: Payback Period | e.g., installing solar panels on municipal buildings |
| Non-Profit Analytics: ; Cost-effectiveness of initiatives | Determine a quantifiable goal ; Analyze the progress, success, and cost of achieving the predetermined goal ; e.g., WGU “online, accelerated, affordable, accredited” ; WGU has not raised tuition rates since 2008 |
| Non-Profit Analytics: Compare | results to that of other non-profits |
| Private sector cost benefit analysis | In private sector, seek Revenues > Cost |
| CPI base period used | 1982-1984 = 100, all else relative to that |
| simple composite index | our score/industry averge * 100 |
| person-time units | cumalative amount of time each person was studied |
| Key Performance Indicators (KPIs) | KPIs are performance measures that organizations use to quantify their level of success |
| Key Performance Indicators (KPIs) | examples include: customer/patient satisfaction, sales increase, employee turnover, return customer rate |
| Management/Employee bonuses | can be tied to KPIs |
| Key Performance Indicators (KPIs) | often follow “SMART” criteria |
| Key Performance Indicators (KPIs): SMART | Specific, Measureable, Attainable, Relevant, Time-bound |
| KPI Dashboards | Visual representation of multiple KPIs |
| KPI Dashboards | include: key areas of focus; often for seeing historical trends; can readily see if organization is meeting its goals |
| KPI Dashboards | use when one chart, graph or piece of data does not provide enough info to make a decision |
| Advantages of KPIs | help company track financial, productivity, etc. goals |
| Advantages of KPIs | data-driven results that make it easier to quantify performance |
| Advantages of KPIs | can be used as a tool across an entire organization |
| Advantages of KPIs | internal benchmarking |
| Disadvantages of KPIs | can be expensive and requires ongoing maintenance |
| Disadvantages of KPIs | small KPI changes that are not statistically significant might be mistakenly viewed as meaningful |
| Disadvantages of KPIs | might lead to focus on short-term, rather than long-term, gains |
| A Balanced Scorecard | measures an organization’s performance on a balanced mix of financial and nonfinancial measures |
| balanced scorecard: variance | actual performance - target performance |
| Balaned Scorecard Measures 4 perspectives | Financial 2. Customer 3. Internal business processes 4. Innovation and learning |
| Balaned Scorecard Measures goal: | positive impact on the company’s long-term performance |
| Balances scorecards include: | mission/vision, objectives, measures, targets & initiatives |
| Balanced Scorecards Advantages | better organization alignment |
| Balanced Scorecards Advantages | better communication |
| Balanced Scorecards Advantages | links operations with company strategy |
| Balanced Scorecards Advantages | emphasizes strategy and organizational results |
| Balanced Scorecards Disadvantages | requires time & effort to develop a meaningful scorecard |
| Balanced Scorecards Disadvantages | challenges for cross-company adoption |
| Balanced Scorecards Disadvantages | may not encourage desired behavior changes |
| Net Promoter Score (NPS®) | considered a measure of customer loyalty |
| Net Promoter Score (NPS®) | developed by Fred Reichheld (Bain & Company) in 2003 |
| Calculate NPS | 1. Ask respondents a single, 11-point (0 to 10) question… How likely would you recommend this product or service to a friend? |
| Calculate NPS | 2. Categorize respondents into 3 groups: Detractors, Passives & Promoters |
| Calculate NPS | 3. NPS = % Promoters - % Detractors |
| Should you proceed NPS | only if your NPS is > or = to industry NPS |
| NPS Advantages | easily understood metric |
| NPS Advantages | easy to calculate and monitor |
| NPS Advantages | ; can benchmark against industry leaders |
| NSP Disadvantages | doesn’t consider in-depth customer perception data |
| NSP Disadvantages | may fail to predict loyalty behaviors |
| NSP Disadvantages | doesn’t address specific areas of dissatisfaction (or satisfaction) |
| NSP Disadvantages | “likelihood to recommend” is no better a question (and may even be worse) than “overall satisfaction” or “overall liking” |
| NSP Disadvantages | 11-point scale may have low predictive validity |
| What type of graphical option could one use to show a visual summary of table data? | bar chart |
| What statistical test could be used to determine if there are significant differences in incomes among the 3 towns? | ANOVA |
| z-TEST | 1 sample mean & 1 population mean, known population standard deviation |
| 1 sample t-test | 1 sample mean & 1 population mean, unknown population standard deviation, known sample standard deviation |
| independent t-test | 2 sample means, unrelated samples |
| paried samples t-test | 2 sample means, related samples |
| 3+ sample means | ANOVA |
| What type of graphical option could one use to show a visual depiction of data showing the year, quarter & number of housing starts? | line (trend) graph |
| What analytic approach should be used to determine if there is a linear trend in housing starts over time? | regression/correlation |
| What analytic approach should be used to forecast quarterly housing starts for the next two years? | regression |
| chi square | need to know relationship between variables, both nominal (frequency counts) |
| correlation | need to know relationship between variables, both interval or ratio (measuring something), relationship |
| regression (1 independent variable) or mutliple regression (2+ independent variable) | need to know relationship between variables, both interval or ratio (measuring something), predition/trend |
| logic regression | need to know relationship between variables, binary dependent variable & independent variables: interval or ratio (measuring something) |
| What type of analysis would you use to compare 3 different compounds that need to be mixed together to create concrete with constraints? | linear programming |
| How might one display this data on county firetrucks for possible sale including town, mileage, age in years & selling price? | scatter plots |
| How could one summarize the relationship between Selling Price and either Mileage or Age? | Correlations |
| What analytical approach would be used to determine if Selling Price can be predicted from Mileage and/or Age? | Multiple Regression |
| What analytical approach would be used to determine the number of units Trinity would need to sell to cover its costs? | Break-Even Analysis |
| How would you visually depict break-even analysis results | Line Graph (Break-Even Chart) |
| breakeven line graph | anything on the left of the break even point between revenue and cost are losses |
| breakeven line graph | anything on the right of the break even point between revenue and cost are profits |
| What analytic approach should be used when calculating whether montly labor costs on a new project experiences monthly cost over-runs | determination of z-scores |
| What measues should be calculated when tracking whether montly labor costs on a new project experiences monthly cost over-runs | upper & lower control limits |
| What type of visual display should be used to show whether there are any monthly cost over-runs on a chart showing the month and costs for a project? | statistical process control chart |
| NPS Promoter | 9 -10 |
| NPS Passives | 7-8 |
| NPS detractors | 7 or less |
| can you have a negative net promotor score? | yes |
| NPS | doesn't explain why they would or would not recommend |
| The key focus of the analytics | is to predict trends using quantitative data |
| use data | to inform you of your options and assist you in making decisions. |
| Nominal data | is categorical. It has no numeric value. Ex. Meatball, veggie, and cheese pizzas. These are labels. They can’t be added or subtracted. |
| Ordinal data | is ranked, but doesn’t have a specific value. Small, Medium, and large are examples |
| Interval data | is numeric. You can add and subtract it. It has a sequential value. Each value is equally spaced from the previous value. You serve 16oz, 20 oz and 24 oz soft drinks. These are equally spaced apart. |
| Ratio data | Your sales per day are ratio data. 10 sales of $12.99 a piece = $129.99. The value has a true value from zero. |
| outlier | Last week you were closed for two days for renovations. Sales were a zero for those two days. |
| when outliers are present | Do you include the two zeros when calculate the average sales for the week? Do it both ways so you know how it’s effecting your bottom line. |
| Random | happens just once and will not repeat over time. If you’re trying to find average delivery times and one delivery was effected by a four hour Chicago traffic delay |
| Systematic | not by chance & repeats. Frank has nursed the fuel injector on his car for the past 6 months. It breaks down 1out of every 20 deliveries he makes. |
| omission error | the driver didn’t clock in or out for his delivery. That data will not be included in your study and it’s relevant |
| distorted | A data set with an omission error |
| out of range error. | Your range is 0 – 15 miles. Delivery 14 and 16 are not in the scope of what you are measuring because they are too far |
| reduce errors | a survey your customer takes via text only allows them to select responses from a list. That way they can’t type anything in wrong. |
| treatment | You are treating the pan three different ways with the three amounts of oil. |
| blind study | The manager doesn’t know the treatment of each pizza. You do. |
| double-blind study | Let’s say you had someone else prepare the pizzas, so you don’t know which one is which. You set out the pizzas for your staff and record the results of each of their taste tests. |
| PDCA: Plan | List every reason a shipment could be delivered to the wrong warehouse. Make a list of possible solutions and select the one you want to try. |
| check sheet | collect when an error occurs for each shipping location. Tally sheet |
| cause and effect diagram | brainstorm the issues casing the shipping error. |
| Pareto | shows us which “cause” happens the most because sort it highest to lowest order |
| Pareto | a bar graph with categories shown in descending order by frequency. |
| PDCA: Do | Implement the solution on a very small scale. We have one clerk trialing a new data entry format when she enters the orders in the computer. |
| PDCA: Do | We should create a flow chart to clearly define the steps in the new process for the clerk. |
| PDCA: Check | Have we had greater success with program we are piloting with the one clerk? Yes. |
| PDCA: Check | To know if we’re getting better, we look at our performance over time. In Six weeks let’s see if shipping errors decreased. |
| run chart | track errors for the last 6 weeks |
| PDCA: Act | We accept the trial method as our new process and implement it with all of the shipping clerks. |
| Plan | nothing’s been done. |
| Check | we are measuring what’s been tried |
| control chart | tells if we stayed within limits we set. (54 – 60 inches) |
| Ishikawa | claims quality tools can solve 90-95% of quality problems |
| SIPOC diagram | list all of the elements that can influence a process before it starts |
| Six Sigma | When we want to correct a part of the process, we can use this as a problem solving method. |
| histogram | showing the number of returned items for each year from 2012-2016. |
| histogram | used rather than a bar chart because our x-axis values are numeric (different years in this case) rather than categories. |
| scatterplot | For each of the 25 manufacturing employees, a point is plotted for their number of overtime hours and the number of errors made in the past 6 months. |
Created by:
seamonkeyink