My research topic is ''Role of classroom management in modifying students behaviours''. I made a questionnaire on likert scale with 5 options from strongly agree to strongly disagree. I collected data from 65 respondents and entered data in spss. I coded the questionnaire response data. Now please guide which type of test or analysis I can run for this data. I analyze descritive data fequencies which out put I have got. Redt I dont know which analysis will be accurate to summarise this data and get results
Depending on the study objectives, if you only have a single-variable survey, as you mention it was created, you should evaluate the structure of your survey through factor analysis and reliability as part of the adaptation for use in your context and then perform the relevant analyses according to your study objectives.
There is an important difference between individual Likert-scored items and scales that combine multiple items into a new variable. If your goal is to create scales, I have collected a set of resources on this topic:
https://www.researchgate.net/post/How_can_one_help_people_create_simple_scales_from_Likert-scored_items
Hello Shoukat,
Both Cristian Ramos-Vera and David Morgan have raised some important points (to wit: What are your research questions? and Are you attempting to use individual items as variables, or combining responses from multiple items into a summated or averaged score?)
To this, I'll add one more question for you: What variables are you attempting to measure/quantify or use in your research?
If you could amplify on these points, I think you'll be far more likely to get guidance that will help you to proceed.
Good luck with your work.
SAV file is attached which is filled with all relevant data. If any one can please check that. Objective of the research is, ''Role of classroom management in modifying student behaviours''.
Shoukat Khan, unfortunately, I can't open a .SAV document. However, it would be helpful if you answered the questions asked above. In order for me to help you initially, I would also like to know how many items you had in your scale and whether you anticipated it having only one "factor" or more than one.
Simple descriptives appear to be the way to go in your case. You could report the frequencies of responses to the questionnaire items. Also you can calculate the summary measures of mean(SD), median(IQR).
There is no detail given in the question if there are any characteristics of the sample basing on which you would like to check the differences in questionnaire responses (for each question). If you have such objectives, you can consider Mann-Whitney for a dichotomous independent variable, Kruskal Wallis H for a multichotomous independent variable. However, sample size of 65 is a concern.
Finally, deriving a scale score for the five questions is not warranted in your context as the choice of questions, content relevance, consistency between items, the structural validity of the items as a unit is not determined. Even if you wish to do it now, it may appear far fetched. Try and integrate all the principles involved in scale development in your next project if you wish to obtain a scale score.
Good luck.
Determining whether a scale score is appropriate is a matter of assessing the extent to which the proposed items all measure the same construct, so that they are all strongly intercorrelated.
At first, you can sum all 5 point Likert questions to obtain total sores. Then, you can perform descriptive statistics like means, median, standard deviations, etc.
- If you want to see group differences with socio-demographic or other variables, you can perform t-tests/one way ANOVA tests.
- Finally, you can perform the hierarchical regression analysis to determine significant predictors.
- Badges
- Science topic