I used 5 point Likert scale Never,Rarely,Sometimes,Often,Very Often,in order to measure the women's harassment experience,.what is good analysis by using these category,Never is saying have not experienced at all, whereas rest of all are showing experience but its frequency may vary,
In my opinion, it depends on the Design considerations.
The data analysis, decision for Likert items should be made at the questionnaire development stage.
• If you have a series of individual questions that have Likert response options for your participants to answer . Then analyze them as Likert-type items i.e. Modes, medians, and frequencies.
• If you have a series of Likert-type questions that when combined describe a personality trait or attitude . Use means and standard deviations to describe the scale. 1
the first thing I can say is: it's ordinal and not interval. you have to finde statistical methods for ordinal measures. a lot of standard methods (ANOVA, linear regression, pearson correlation, etc.) require at least interval scales.
Don't use a likert scale variable as an interval dependent variable, unless you are willing to assume that the distances between ordered items is precise amen dent icecap across items.
For ordinal dependent variables, you can use ordinal logistic regression.
http://www.ats.ucla.edu/stat/stata/dae/ologit.htm
As opposed to purists who have responded so far, the most common way to create a scale from a set of Likert scale items is simply to add them up. The metaphor here is like a bundle of sticks, where a single measure may be weak, but the full set is much stronger.
To assess how well the items go together to form a single scale, you need to use Cronbach's alpha which is available in almost all statistical packages (in SPSS, you will find it under Analyses: Scale: Reliabilities.
Of course, you need to be sure that other researchers in your field use this method of scale construction -- which is quite common throughout the social sciences. The easiest way to tell is by seeing if people do indeed report Cronbach's alpha as a measure of reliability, and then when you see that, check whether they are adding together Likert type items.
I agree that works best if you have multiple items and can create an index. If you only have one ordinal variable, then I'd recommend ordinal logistic regression.
Thank you very much all for your detailed input, concern and Time. Your effort is much appreciated
In my opinion, it depends on the Design considerations.
The data analysis, decision for Likert items should be made at the questionnaire development stage.
• If you have a series of individual questions that have Likert response options for your participants to answer . Then analyze them as Likert-type items i.e. Modes, medians, and frequencies.
• If you have a series of Likert-type questions that when combined describe a personality trait or attitude . Use means and standard deviations to describe the scale. 1
I agree with Sana that the statistical analysis would have been part of the design stage or hypothesis formulation. None the less, from the data (and variable) you presented it would seem you have to start with descriptive measures such as frequency count and ranking.
Hi everyone. I am sorry to interrupt this discussion but I was wondering if i could get some expert advice on how to analyse my data.
I am trying to run the statistical analysis for a cross cultural study comprising of two country samples. I have 5 variables, one mediator(perceived value) and 3 moderators (subjective wellbeing, healthy lifestyle, ethical identity), IV (conspicuous consumption), DV(customer loyalty) all measured on a 7 point likert scale. I was wondering if 3 moderators is too many and what would be the statistical tests I should perform for my study.
It sounds like you have continuous moderator variables, which can be tricky to interpret, so I suggest you do a search on "moderation with continuous variables." In general, it would be possible to examine each moderator individually, but if you want to look at higher-order interactions among the moderators, this would be very difficult with continuous variables.
Thank you! David L Morgan to analyse the different cultures between 4-5 countries, would it be ideal to perform an ANOVA?
- Badges
- Science topic