Factors of affecting choice of statistical test

Question 1

3 things to consider

Answer 1

1. Is it a test of DIFFERENCE or association/ correlation (so consider aims and purposes, if you see word relationship it’s a test of correlation)
2. experimental design?
3. what level of measurement is used

Question 2

levels of measurement are

Answer 2

NOMINAL DATA- Data in categories for example the number of boys who conformed, 1 item in 1 category (think of nominal- named, so can be gender, hair colour etc)

ORDINAL DATA- ordered in some sort of way and is subjective eg let’s say for a race, we have 1st 2nd 3rd etc, there’s no standardised difference in time between these scores (think of order or SCALES)

INVERTAL DATA- based on numerical scales eg time temperature (so the difference between points is standardised and hold much more meaning)

Question 3

mnemonic for learning tests to be used

Answer 3

Carrots Should Come Mashed With Swede Under Roast Potatoes

Chi- squared (IG), Sign test (MP OR RM), Chi- squared (association), Mann Whitney (IG), Wilcoxon (MP OR RM) Spearman’s Rho, Unrelated t-test(IG) , Related t-test (MP OR RM), Pearson’s r*
(* means parametric data which is under interval data)

Question 4

what tests are used under association or correlation

Answer 4

Chi squared (association)
Spearman’s Rho
Pearson’s r*

Question 5

what does Pearson’s R test, Related, unrelated t test and chi squared use to check critical value

Answer 5

Pearson- df= number of pps - 2
Related t test= df= Na + Nb - 2
Unrelated t test= df=N-1
Chi squared= df= (rows-1) x (columns-1)