Hey researchers! I am currently doing a research about to what extent in which A has accelerated the inclusiveness of the payment system in my country. Down below are a few examples of the questions in my survey.
BEFORE
Please rate the convenience of payments before using A system:
☐ 1. Much less convenient ☐ 2. Less convenient ☐ 3. Neutral ☐ 4. More convenient ☐ 5. Much more convenient
Please rate the speed of payments before using A system:
☐ 1. Very slow ☐ 2. Slow ☐ 3. Average ☐ 4. Fast ☐ 5. Very fast
Please rate the cost of fund transfers before using A system:
☐ 1. Very inexpensive ☐ 2. Inexpensive ☐ 3. Average ☐ 4. Expensive ☐ 5. Very expensive
AFTER
Please rate the convenience of payments after using A system:
☐ 1. Much less convenient ☐ 2. Less convenient ☐ 3. Neutral ☐ 4. More convenient ☐ 5. Much more convenient
Please rate the speed of payments after using A system:
☐ 1. Very slow ☐ 2. Slow ☐ 3. Average ☐ 4. Fast ☐ 5. Very fast
Please rate the cost of fund transfers after using A system:
☐ 1. Very inexpensive ☐ 2. Inexpensive ☐ 3. Average ☐ 4. Expensive ☐ 5. Very expensive
Do you think with these questions is it possible to do apply paired sample t-test? If not, what is the better alternative you can suggest?
Hello Kong Solydeth. The items you showed do not look (to me, at least) as if they sum to give a total score. Rather, it looks (to me) as if you want to compare before & after for each individual item. If so, that makes use of a paired t-test more difficult to defend. I suggest that you take a look at:
Agresti, A. (1983). Testing marginal homogeneity for ordinal categorical variables. Biometrics, 505-510.
https://users.stat.ufl.edu/~aa/articles/agresti_1983.pdf
See Section 2 (Ordinal Strategies). HTH.
PS- You may find other helpful resources on Dr. Agresti's web-page:
- https://users.stat.ufl.edu/~aa/
I concur with Dr. Yahaya, though it really depends on how you visualize your data. For ordinal variables, McNemar is an acceptable choice given the paired nature of the variables. Analyzing the values numerically would necessitate the use of either Wilcoxon signed-rank or a paired sample t-test, though the latter method is a tad bit controversial. Hope this helps.
Thank you so much, everyone, for your replies. Although I am in my second year of my master's program, these kinds of terms and analyses are still very new to me. However, your replies have led me to learn and research these terms and analyses. It was stressful, but I am very happy because I feel like I have gained a better understanding. I am sure there will be more things that I have to rack my brain for in the future, but as long as it helps me improve my research, I am willing to do it.
- Badges
- Science topic
- Similar topics
- Social Psychology
- Attitude