click below
click below
Normal Size Small Size show me how
LECOM Biostatistics
LECOM Biostatistics Freeman
Question | Answer |
---|---|
Seven Core Competencies | OPP, Pt Care, Medical Knowledge, Practice-Based Learning and improvement, Interpersonal and Communication Skills, Professionalism, Systems-based Practice |
Quantitative | each value is a number that represents an amount or count *interval/ratio |
Qualitative | each value is a word or code that represents a class or category *nominal/ordinal |
Mean | sum of all observations divided by N (number of total observations) |
This MCT is most sensitive to outliers | Mean |
Median | middle value when observations (ordinal) are ranked in order |
Use this MCT when data is skewed | Median |
Mode | value most frequently observed |
When can only the Mode be used | with nominal data |
Find variance by | square individual deviation scores then average them |
Standard Deviation in relation to the variance | Square root of the variance |
The normal curve is | symmetrical and has mean=median=mode |
Skewedness is named + or – based on | where the outlier is |
Kurtosis describes the | peakness or flatness of a curve |
Leptokurtic curve is | thin & peaked |
Platykurtic curve is | flat and wide |
Causality must be determined because | statistical correlation does not always mean causation |
r=correlation coeff. Which ranges from | -1 to 1 |
Positive or negative reflects | slope of the line |
The closer to 1 the coeff. Is the | more correlated the data is |
R*2 (squared) shows | amount of variation in the dependent variable explained by the independent variable |
Total number of cases of disease or injury in a population at a particular point in time or during a specific time period | prevalence |
Measure of NEW cases in a population over a specific time period | incidence |
Those who have the disease and tested positive divided by the total number of those who have the disease | sensitivity |
The ability of a test to correctly identify those who do not have a disease | specificity |
Probability that the person tests negative if they do not have the disease | specificity |
Probability that a person will test positive if they have the disease | sensitivity |
Likelihood that a positive test is predictive of having the disease | PPV |
PPV = | True positives divided by sum of all positive tests |
Likelihood that a negative test is predictive of being disease free | NPV |
Likelihood Ratio of Positive test means | how much more likely is an infected Pt to test + than a disease free Pt |
Likelihood Ratio of Positive test calculation | Sensitivity/(1-Specificity) |
Likelihood Ratio of Negative test means | how much more likely is a disease free person to test – than an infected Pt |
Likelihood Ratio of Negative test calculation | (1-sensitivity)/Specificity |
CER | Control Event Rate |
EER | Experimental Event Rate |
AAR | Absolute Risk Reduction |
RRR | Relative Risk Reduction |
AAR = | EER-CER |
RRR = | (EER-CER)/CER |
Number Needed to treat means | number you have to treat to “save” one |
NNT = | 1/AAR (negative value) |
Number Needed to Harm means | number of patients you need to treat to have one with adverse effects |
NNH for control and experiment = | 1/ AAR (positive value) |
Hypothesis testing step 1/6 | State the question |
Hypothesis testing step 2/6 | Formulate Null and Alternative hypothesis |
Hypothesis testing step 3/6 | Establish a decision rule |
Hypothesis testing step 4/6 | Do the research |
Hypothesis testing step 5/6 | make a decision |
Hypothesis testing step 6/6 | Interpret |
Decision rule uses p which means | what probability of being wrong are we willing to tolerate |
Decision rule using p=0.05 means | if p<0.05 reject the null hypothesis |
Type I error (alpha) means | Determined the treatment works but it actually doesn’t |
Type II error (beta) means | Determined the treatment didn’t work when it really does |
Power refers to a test’s ability to | find a difference if one really exits |
Power = | 1-beta |
z-statistic can only be used with | a NORMAL CURVE ie ideal case |
If 1/(EER-CER) > 0 than it is the Number Needed to | Harm |
If 1/(EER-CER) <0 than it is the Number needed to | Treat |