Research Methods and Stats

Question 1

One (two, three) way ANOVA

Answer 1

The one-way anova is used when there is one IV (regardless of the number of levels) and one DV. By extension, the two-way anova is used when there are two IVs (regardless of the number of levels) and one DV; the three way anova is used when there are three IVs and one DV.

used to compare difference between groups. F-ratio

Question 2

Factorial ANOVA

Answer 2

The factorial anova is used when there is more than one IV (e.g, two IVs, three IVs, etc) and one DV.

Question 3

Chi Square

Answer 3

Chi square is used when the dependent variable is nominal or categorical.
Used to analyze the frequency of observations in each category (like how many people prefer each poliitcal candidate) and if that differs from the expected frquency

Question 4

MANOVA

Answer 4

is used when there are multiple dependent variables.

Question 5

Normal Curve (z, T, %, scaled score, Standard Score)

Answer 5

T: mean = 50, SD = 10
Z: mean = 0, SD: 1
%: mean = 50, SD = 34
Scaled score (ss): mean = 10, SD = 3
Standard score (SS): mean = 100, SD = 15

Question 6

ABAB design

Answer 6

The threat of history is best taken care of by an ABAB design in which the person is first measured over time during a baseline phase (A), then treatment is implemented and the person is once again measured repeatedly during the intervention (B). Treatment is then removed, and the person is measured at baseline again (A). Finally, treatment is implemented once again and the person is measured repeatedly (B). Since measurements of treatment effect are obtained at two different times, this design controls for the possibility that some of the treatment effect was confounded by a historical event.

Question 7

Standard error of the measurement vs. standard error of the estimate

Answer 7

SEM: used to create a confidence interval around a person’s obtained score
SEE: Used to create a confidence interval around a predicted criterion score

Question 8

Type 1 error vs. Type 2 error

Answer 8

1: false positive
2: false negative

Question 9

sensitivity vs. specificity

Answer 9

Sensitivity: the percent of people in the validation sample who have the disorder who were acurately IDed by the predictor as having the disorder.

Specificity: the percent of people in the validation sample who DO NOT have the disorder who were acurately IDed by the predictor as NOT having the disorder.

Question 10

cross validation

Answer 10

Sometimes criterion are influenced by chance factors associated with the sample characteristics. important to cross validate a predictor with another sample. Cross validations coeffieicnts tend to skrink

Question 11

Split half reliability

Answer 11

intervnal consistency measure.
split test into two equal parts.
spearman brown prophecy formula used to correct for halfing the number of items.

Question 12

Chronbachs alpha

Answer 12

average degree of inter item consistency.
basically calculates the average reliability of all possible splits.
KR-20 variation can be used if the test is dochotomous

Question 13

Kappa coefficient

Answer 13

inter-rater reliability

Question 14

central limit theorem

Answer 14

as sample size increase, the sampling distribution of the mean approaches a normal distribution

Question 15

standard error of the mean

Answer 15

measure of variability that is due to the effects of sampling error (SEM = SD/ square root of N)

Question 16

Null hypothesis

Answer 16

independent variable does NOT have an effect on the dependent variable. Alternative hypothesis is the oppostie

Question 17

Alpha

Answer 17

significance. basically measures the size of the rejection region. ussually set at .05 or .01.

Question 18

Beta

Answer 18

the probability of making a type 2 error (false negative). Power equals 1-Beta. Power is your ability to reject a false negative.

Question 19

Increase Power

Answer 19

increase sample size
increase effects of IV
minimize error
one tailed is more powerful
parametric tests are more powerful (t-test, ANOVA)

Question 20

ANCOVA

Answer 20

version of ANOVA that is used to increase efficiency of analysis by removing varibaility due to an extraneous variable

Question 21

t-test

Answer 21

used to compare two means

Question 22

Multiple regression

Answer 22

used to predict a score on a continuous criterion based on performance on two or more continuous or discrete predictors. Output is R. you want low correlation between predictors (multicolinearity) and high correlation with criterion.

Question 23

Cohens D

Answer 23

effect size. measure of the difference between 2 groups in terms of SDs. small = 0.2; med = 0.5; large = 0.8

Question 24

Eta squared

Answer 24

percent of the variance in the outcome that is accounted for by variance in the treatment.