Page 1 of 1

non-integr figures after likert data aggregation

Posted: Tue Feb 07, 2017 9:53 pm
by alizol
Dear Smartpls Experts
What Should I do in my csv file if I have more than one respondent for measures and I need to aggregate different responses as if I have only one respondent? If I use mean for aggregation of Likert data of all respondents, I may have some non-integer results.

1-Am I allowed to enter the non-integer figure in SPSS to be imported to Smartpls? For example after aggregation, mean of answers for the first indicator may be 1.67.SO should I round the no to 2?
2-Doesn't this method for aggregation of different respondent lead to incorrect model estimates?

Re: non-integr figures after likert data aggregation

Posted: Wed Mar 01, 2017 8:12 pm
by cringle
@1: Yes!

@2: That's what you need to decide and justify (to what extent you manipulated the data).

Kind regards
Christian