Two-sample t-test appropriate?

Question

I have two hypotheses in which there is 1 independent variable (2 separate samples). I want to find out if the outcome of each of the 2 dependent variables is affected by the independent variable. Would it be appropriate to conduct a two-sample t-test, once for each dependent variable? Or, would it be appropriate to conduct a one-way ANOVA, even though there are only 2 separate groups?

EX:

Younger participants will have more positive attitudes toward eating pizza than older participants. (hypothesis 1)

&

Older participants will have greater perceptions of stigma associated with eating pizza than younger participants. (hypothesis 2)

The dependent variables are ordinal. A four point Likert scale will be used to score the responses. — A. Iman, Oct 10 '15 at 07:05

score 0 · Answer 1 · edited Jun 03 '19 at 20:46

0

This link has a very good explanation of why you should compare multiple groups with ANOVA instead of multiple t-tests.

I believe it will answer your question, but to summarize, as long as you have two groups and different samples, a t-test would be a good way to go.

edited Jun 03 '19 at 20:46

mkt

11,770
9
51
125

answered Oct 09 '15 at 19:37

Pappu Jha

101

The appropriateness of a t-test depends on how the dependent variables are measured/coded. For example, if they are binary, t-test would not be a good idea. – Sheep Oct 09 '15 at 19:41
The issue here is actually not about multiple groups versus two groups (as in your link) but about multiple dependent variables versus one dependent variable. So it is not about multiple two sample t-tests versus anova, but instead about anova versus manova. – Sextus Empiricus Nov 12 '20 at 15:15

score 0 · Answer 2 · answered Oct 12 '15 at 23:40

0

A test to compare ordinal outcomes in two separate groups would be the Mann-Whitney U test.

https://en.wikipedia.org/wiki/Mann%E2%80%93Whitney_U_test

answered Oct 12 '15 at 23:40

Sheep

504
3
15

Two-sample t-test appropriate?

2 Answers2