I have 378 participants and have 18 questions regarding 5 point Likert scale. The questions are like this
I believe that most of the friends in my age smoke
most unlikely 1.............2...............3................4.................5 most likely
Now I wonder that how can I interpret this question. Should I interpret every single question or in combined?
It all depends on the direction of your hypothesis. If you intend to measure positive attitudes concerning tobacco use , then the scale points : strongly agree, agree, undecided, disagree , strongly disagree should be rated as 5, 4, 3, 2, 1 . Otherwise, the ratings must be reversed.
In my opinion, each point on the likert scale must be defined beforehand. If not, it may be misinterpreted by the respondent.
Interpreting each question gives you more information, but it gets lost in the overall result. One solution is to group 1-2, 3, 4-5. However, it would only be appropriate if the sense of each group is inverse. That is 1-2 negative, 3 neutral, 4-5 positive. In this case, it would be possible to represent the pooled value (general vision) and for each point. Thus, two answers where the pooled value (example: 1 -2 have the same percentage, 50%) could have different weight depending on whether the majority is in 1 or in 2. Example: Alternative A: 1 = 45%; 2 = 5%. Alternative B: 1=25%; 2=25%).
The problem is that since the intermediate values are not defined, it could be interpreted on a positive linear scale (perhaps this is how I would interpret the question).
It will depend on your research question. What are you trying to find out from this data? Why did you collect this data in the first place? Can you compare attitudes according to age, gender, location? - Analysis is dictated by the RQ.
Hope this helps :)
Thank you both of you and I agree with Hector D De Paz and Reza Biria
The mean score of each item should be compared with the average score of
the choices (i.e. 3.00). Hence, if the mean score of a item was less than 3.00, the
respondents tended to disagree with that statement. On the contrary, a mean score above 3.00
indicated the respondents’ inclination to concur or agree with that item.
Thank you for wonderful information, Ehsan Namaziandost however, should we interpret the each item or combined score as per the average score?I would be grateful if you further elaborate on it
Dear Bahadur
after calculating each item mean, you can run a one-sample t test to see if the positive or negative attitudes reached statistical significance.
Dear Bahadur after calculating each item mean, you can run a one-sample t test to see if the positive or negative attitudes reached statistical significance.
… @ Bhakta Bahadur K.C.
- Badges
- Science topic
- Similar topics
- Social Psychology
- Attitude