Stats Final Exam

descriptive statistics

organize/summarize data (median, mode, mean)

Inferential statistics

utilize sample data to draw conclusions about a population

parameter

number generated from everyone/everything in a population

statistic

number generated from everyone/everything in a sample

categorical data

categories and frequencies

measurement data

measuring objects or events (time, length, weight, 0-10 scale, scores, etc)

nominal data

numbers are only labels, they mean nothing (political affiliation, age, gender, religion, etc)

ordinal data

variables that are ranked/ordered; differences between scores not meaningful (senority)

interval data

differences in scores are meaningful; no true zero point (rank from 1-10, temperature)

ratio data

existence of a true zero point (length, volume, time, mean, likert scale)

discrete variable

few possible variables; all nominal; obtained through counting (gender, religion, # of cars, etc)

continuous variable

can take on any point value between the lowest and highest points on the scale; obtained through measuring (time, weight, GPA)

independent variable

the groups were comparing, the predictor

dependent variable

what we are comparing the groups on, the criterion

normal distribution

highest point in middle

positively skewed distribution

tail on high end

negatively skewed distribution

tail on low end

mode

the most frequently occurring score in your data set. (best for nominal data or when you need to know what happens the most)

median

50th percentile; (n+1)/2 (best for when you don’t want the outliers to count")

mean

the average (best parameter estimator, use in every other situation)

variability

how different your scores are from one another

range

high score - low score

interquartile range

range of the middle 50% of observations (find median, then find median of both sides

standard deviation interpretation

on average, scores are falling __ units from the mean

z-score

standardized scores; mean is always 0, standard deviation always 1

probability

the likelihood that an event will occur

elementary event

can’t be broken ito smaller events

compound event

the probability of multiple events together

independent event

if more than 1 event, the occurrence of one thing happening doesn’t change the probability of the other

mutually exclusive event

when the occurance of one outcome eliminates the occurrence of another

exhaustive event

a set of events where all possibly outcomes have occurred

sample with replacement

sampling in which the item drawn on trial n is replaced before the next draw

the additive law of probability

you have one outcome or one action

multiplicative law of probability

there are multiple outcomes and we want to know the probability of all of them

joint probability

the probability of the co-occurrence of two or more events

conditional probability

probability that one event will occur given the occurrence of some event

interpretation for single score z-test

we found evidence to say/we didn’t find enough evidence to say…

confidence limit interpretation

I’m __% confident the population mean says…

sampling distribution of the mean

all possible samples

sampling error

chance the difference between X & M is due to _____.

Central Limit Theorem

Standard distribution of the mean = M, SDM, SDM always normally distributed with at least n=30

Factors that affect likelihood of rejecting H0

1. X bar - M gets larger, so does t

2. S goes up, t goes down

3. n goes up, t goes down

4. alpha goes up, change of reject up

5. 1-tail more likely to reject

6. increase st dev, decrease change of reject null

normality

two populations have normal distributions

homogeneity of variance

two population variance are the same

independence of observations

all scores independent of one another

power

if null is false, power is your probability of correctly rejecting the null; probability of finding evidence that you are right

effect size

how big is the effect in the populations?

Interpret power

if the null is false, there is a __% change of correctly rejecting the null

how to increase power

increase n, increase alpha, increase distance between means, one tailed, decrease standard deviation

matched samples

different people but the scores are related

order effect

order in which treatment was administered effects performance

carry over effect

previous trials effect participants performance on subsequent trials

magnitude of effect

(IV) accounts for __% of the variability in (DV).

error of variability

variability within groups; can’t explain