Session 4b

Question 1

Prior requirements/assumptions of one way ANOVA

Answer 1

The population distribution of the DV is NORMAL within each group The variance of the population distributions are equal for each group (HOMOGENEITY OF VARIANCE ASSUMPTION)
Independence of observations

Question 2

The assumptions about normality and equal variances are assumptions about what

Answer 2

the population

Question 3

The assumptions about normality and equal variances are assumptions about the population
Usually the best we can do is examine the what

Answer 3

sample for evidence about whether these assumptions hold

Question 4

what re the methods for assessing normality

Answer 4

Descriptive and Inferential Statistics:
Tests for skewness
K-S and Shapiro-Wilk tests

Visual methods:
Histograms
Normal Quantile (Q-Q) Plot

Question 5

what is Skewness

Answer 5

represents symmetry and whether the distribution has a long tail in one direction

Question 6

Look at descriptive statistics

Skewness should be what

Answer 6

≈= 0

Question 7

> 0 skewness indicates what

Answer 7

positive/right skew

Question 8

< 0 skewness indicates what

Answer 8

negative/left skew

Question 9

what does the The Shapiro-Wilk test do

Answer 9

Compares sample scores to a set of scores generated from a normal distribution with the sample mean and standard deviation

Question 10

what is the Limitation of the normality tests

Answer 10

It is easy to find significant results (reject null hypothesis that data is normal) when sample size is large

Question 11

what should you do inanition to the normality tests

Answer 11

plot the data

Question 12

Separate histograms for each group to assess normality: what to look for

Answer 12

Look for obvious signs of nonnormality

Does not have to be perfect, just roughly symmetric

Question 13

what is the problem with constructing plots

Answer 13

Can be difficult to assess visually

Question 14

how to Evaluate a normal quantile plot (or normal probability plot) (the graphs)

Answer 14

Sort observations from smallest to largest
Calculate z-scores for the sorted observations
Plot the observations against the corresponding z-scores
If the data are close to normal, then the points will like close to a straight line

Question 15

Serious violation of Assessing homogeneity of variance tends to inflate what

Answer 15

the observed value of the F statistic

aka Too many rejections of H0 (high Type I error)

Question 16

why can we not just do the Fmax test for homogeneity of variance

Answer 16

Easy to compute, but assumes that each group has an equal number of observations

Question 17

what does the Levene’s test test

Answer 17

homogeneity if variance; Tests the null hypothesis that the population variances are equal

Question 18

what does the levee’s test measure

Answer 18

It measures how much each score deviates from its group mean

Question 19

for levene’s test, It is very easy to obtain significant results when what

Answer 19

the sample size is large

Question 20

If normality and homogeneity of variances is not met. . what should we do

Answer 20

A data transformation may be useful

Question 21

A data transformation may be useful to. .

Answer 21

make data less skewed

make heterogeneous variances more homogeneous

Question 22

If data transformation does not help meet assumptions, what should we do

Answer 22

consider nonparametric tests

Question 23

what is the nonparametric tests

Answer 23

Kruskal-Wallis ANOVA

Question 24

Assessing independence of observations (an assumption of one way anova) why bother doing this

Answer 24

Knowing the value of one observations gives no clue as to that of other observations

Question 25

Independent observations:
Knowing the value of one observations gives no clue as to that of other observations

This is one of the most crucial assumptions underlying what test

Answer 25

the F test

Question 26

It is difficult to predict how bad the F test will be if what is violated

Answer 26

independence of observation

Question 27

is there an easy way to fix the F test when this assumption is violated

Answer 27

no

Question 28

is there easy test for non-independence

Answer 28

no!

Question 29

what can give some clues regarding this assumption (independence of observations)

Answer 29

Knowledge of how data was collected

Example: suppose instead of sampling individuals randomly from a population, we sample pairs: a participant and their best friend The responses of pairs may be similar to each other and NOT independent

Question 30

Any solutions? (for independece of observations)

Answer 30

Experiments should be carefully designed to avoid non-indepndent observations
Random sampling and random assignment should be applied as much as possible

Question 31

what are the steps for One way anova

Answer 31

Assure independence of observations
Check normality and equal variance assumptions
Create ANOVA summary table
If H0 is rejected, conduct multiple comparisons for pairs of means as necessary/desired