🔘 Research Methods: Stastical Tests

Question 1

2 key reasons for stastical tests

Answer 1

Help us determine if the results of a study are due to chance or if there is a real, significant effect

Allow us to accept or reject the null hypothesis

Question 2

3 key criteria for selecting a stastical test

Answer 2

1. Level of measurement (nominal ordinal interval)
2. Experimental design (independent groups, repeated measures)
3. Hypothesis, test of difference or association

Question 3

Spearman’s Rho

Answer 3

Rho
COS - Correlational Ordinal Spearman’s

Test of: correlation
Used to determine whether relationship between 2 co variables is significant or not
Correlation 0 is no correlation and +1.0 means there would be a perfect positive correlation below 0 is negative correlation

When to Use?
- hypothesis predicts a correlation between 2 co variables
- 2 sets of data are pairs of scores from one person, meaning data is ORDINAL

Question 4

Pearson’s r

Answer 4

r
CIP - correlational interval pearsons

Test of: correlation
Used to determine whether the relationship between 2 co variables is significant or not
Each variable must be at interval level of data

When to use?
When hypothesis predicts a correlation between 2 co variables
2 sets of data from one person
INTERVAL level of data

Degrees of freedom for Pearsons r =
Number of pairs of data - 2 (N-2)

Question 5

Mann Whitney

Answer 5

U
MIDO - Mann Whitney independent difference ordinal

Used when hypothesis predicts a difference between 2 sets of data
2 sets of data are from separate groups ( 2 conditions) independent group design
Data is ORDINAL

Question 6

Unrelated t test

Answer 6

T (small DF-2)
UNIITED - unrelated interval independent groups t test

Test of difference
Used to determine if there is a significant difference between the means of 2 unrelated groups
(Unrelated meaning participants in each group completely different; independent group design)

Why UtT
Difference between groups
Independent group design
Interval data

Degrees of freedom :
Df = (n1+n2) -2
N1 =. NUmber of participants in group 1
N2 = number of participants in group 2

Question 7

Wilcoxon Signed Ranks test

Answer 7

WORD - wilcoxon ordinal repeated measures difference

Test of difference, difference between 2 conditions (not a relationship or correlation)
Repeated measures or matched pairs design: same participants in both conditions (RM) or matched pairs when participants matched on a variable and one is assigned to each condition
Ordinal data

Key steps:
Calculate differences
Rank absolute differences
Sum positive and negative ranks
Find the smaller sum (W value)
Compare W to critical value (W must be less than or equal to critical value to be significant )

Question 8

Related T-Test

Answer 8

DIRRT - difference interval repeated measures related t test

When?
Test of difference
Related design - repeated measures or matched pairs
Interval or ratio data

Degrees of freedom =
Df = N-1

For related t test, calculated t value must be equal to or greater than the critical value to be significant

Question 9

Chi Square

Answer 9

X (power of 2
CANDI - chi square association nominal difference independent groups

Test of association / difference
Used to determine whether association / difference relationship between 2 categories is significant or not

When to use?
Hypothesis predicts a difference or association between 2 categories
2 sets of data are pair of scores in the form of frequencies from different people
Data are NOMINAL

Question 10

Remembering statistical tests

Answer 10

Sign = SNoRD
Searmans = COS
pearsons =. CIP
Wilcoxon = WORD
Man Whitney = MIDO
related = DIRRT
unrelated = UNIITED
Chi square = CANDI

Question 11

Remember the test

Answer 11

SALLY SAYS PERFORM WORK MORE RUC