I 'm conducting a research paper on students' attitudes and motivational orientations (instrumental, integrative and instrinsic) toward learning English language.
One of the aims of my research study is to see if there are any differences in students' attitudes and motivational orientations based on their gender.
I used a 6-point Likert scale questionnaire (1- strongly disagree to 6-- strongly agree) for data collection on attitudes and motivational orientations.
Thanks in advance
It depends on exactly what you want to find out. Most people consider a Likert-scale to be continuous, but some treat it as categorical. For the continuous case:
1. Mean differences between males and females in attitude/motivation. This could be simple t-tests for each variable (attitude, motivation, etc.)
2. Are there gender differences in the correlations between attitudes and motivation. This could be a z-test to compare independent samples, i.e., compare males and female correlations.
3. Does gender moderate the attitude-motivation relationship. This could be moderated multiple regression with gender as the moderator.
Regress the questionnaire responses on gender using ordinal regression.
(If you want to examine all the different items in the same model, you would have to either use an item response theory approach, or use mixed-effects ordinal regression [to examine the interaction between gender and item, while accounting for the random effect of participants].)
It depends on exactly what you want to find out. Most people consider a Likert-scale to be continuous, but some treat it as categorical. For the continuous case:
1. Mean differences between males and females in attitude/motivation. This could be simple t-tests for each variable (attitude, motivation, etc.)
2. Are there gender differences in the correlations between attitudes and motivation. This could be a z-test to compare independent samples, i.e., compare males and female correlations.
3. Does gender moderate the attitude-motivation relationship. This could be moderated multiple regression with gender as the moderator.
Thank you all for your help. Paul E. Spector I want exactly to know who ,males or females, have more positive attitudes toward English than the other and what gender category is more integratively-oriented, instrumentally oriented, intrinsically-oriented to learn English than the other. I want to provide very detailed comparisons between males and females in all the variables. Thank you!
independent sample t test would suffice but u need to take type of scale into consideration.
Dear Ismaile
You have to be clear on the qualitative areas that you are taking up for the study.
It seems you are trying it on 3 aspects, instrumental, integrative and intrinsic.
Firstly, choose the appropriate scaling, in the intstruement, 5 point or to be more precise, since these aspects are qualitative, use 7 point, (and clarify which is more influencing than the other)
1. correlate- which influences the most or overlaps or according to the objecitve.
(correlation coefficient) , pearson's /spearman's etc would be fine.
2. use regression analysis- regression analysis estimates the relationship between two or more variables (since you are using 3 variables), and if you want to know the two different group effect, use LASSO Regression analysis.
3. Take one variable and try to find the level of influence on one another with 3 way ANOVA, may be it will lead to interesting leads.
4. Use SEM (STRUCTURAL EQUATION MODELLING) to test the paths between variables in motivations, and / or use testing on direct/indirect variables.
Will get interesting outcome to interpret , all the best.
Humbly state, it all depends, on the intensity of the tool of analysis, methods are wide open still.
regards
KARTHIKEYAN C
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques