| Answer | |
| independent variables | researcher manipulates/researcher can assign you/ cause (not result) |
| dependent variable | measuring behavior/ result (not cause) |
| correlation | 2 variables that are related |
| population | complete collection of anything regardless of size |
| sample | subset of population/ measure so we can talk about population |
| parameter | numerical summary characteristics of a population
stuff you can measure about a population |
| statistics | numerical summary of characteristics of a sample |
| nominal scale (scale of measurement) | assigns names or label to objects or events
not related numerically just assigned |
| ordinal scale (scale of measurement) | puts objects/events in RANK order
the difference btwn 1&2 doesn't mean anything just the order is meaningful |
| interval (scales of measurement) | equal intervals btwn numbers
rank & equal distance
zero isn't true zero, its a point on a scale |
| ration (scales of measurement) | zero means absence of what your measuring
true zero
absolute 0 degrees K no heat
height 0.00 inches
quantity/time |
| frequency distribution | create a table of the data
gives organization
table shows # of times a given score/ group of score occurs |
| rules of graphing | x= variable being studied
y-axis=should be 3/4 the the length of the x-axis
*always be divided using full range of frequency
*always begin y-axis at zero
*if your using top half, indicate a break |
| bargraph | histogram but bars dont touch |
| histogram | depicts grouped frequency distribution |
| grouped frequency distribution | range
grouped data |
| central tendency | trying to come up with 1 number to describe variables
mode
mean
median |
| mode | most frequently occurring score in a group of scores
x |
| median | score that separates top half of group of scores from bottom half
normal curve-median & mode in same place
*data has to be in order
middle score |
| percentile | score at or below which a given percentage of the scores lie
50th percentile: 50% of score are below that score |
| mean | arythmetic average/ sum of scores divided by number of scores
(sum)(x-score)/N-number |
| mean properties | if you take all score the sum of the difference between each score and mean will be 0
sum of squared differences btwn each score from mean will be smaller than sum of squared diff. of any score
1,2,3,4,5
5-3=2
4-3=1
3-3=0
3-4=-1
3-5=-2 |
| z-score | aka standard scores
the deviation of the raw from the mean in standard deviation units
z=raw score(x)-(mean)/SD |
| probability | the probability of event occurring is proportion of times event would occur if chance of occurrence were infinite |
| statistical hypothesis | hyp- guess about population based on sample results
use probability theory to determine degree of certainty we have abt experimental results |
| 2 rules of probability | 1. each event independent from each other/ stand alone. necessary for simple probability
2. dice or coin dont have memory and cant remember or keep track |
| gamblers phallacy | (wrong belief)
mistaken belief that probability of particular event changes with long strain of some event
i.e. slot machines |
| addition rule | mutually exclusive random events
compute the probability of one OR the other occurring
P(A or B)= PA + PB
2 of spaids or 3 of clubs
1/52 + 1/52 = 2/52 = 1/56 |
| multiplication rule | probability of 2 or more independent events occurring is the product of individual probabilities
P rolling 2 ones (snake eyes)
(1/6) (1/6) = 1/36 |
| non independent events | *conditional probability
drawing card and holding onto it
draw ace and draw another one
(4/52)(3/51)=1/221= .0045= .45%
probability of second dependant on first |
| confidence interval | specific kind of interval estimate
interval estimate of population
can define by how "confident"
compute a way to estimate things about a population |
| interval | range (set) of numbers |
| sampling distribution of mean | distribution of means of many samples
u= mean of population |
| sampling distribution of mean (rules) | 1. if you have infinite #, mean of means would be u
2. the more samples you get, the more the frequency graphs will look like normal curve
3. SD of means= standard error of mean
4. the larger each sample size, the smaller the standard error of mean |
| central limit theorem | the more samples you get, the more the frequency graphs will look like normal curve |
| t-score | how far a group mean is from population mean in standard error units |
| variable | anything that can take on different values or amounts |
