Considering the fact that HBM constructs are based on the Likert scale, is it mandatory to analyse the constructs as Mean and Standard deviation. I am of the opinion that it is better analysed as proportions (percentages). This is because respondents who give neutral answers (3) would tend to have a higher value than those who actually give answers and have a perception (disagree or strongly disagree).
Answers would be appreciated as I am only coming across literature that analyse HBM constructs as mean and standard deviation
Thank you
Dear Olutomi Yewande Sodipo
Since the model incoparetes a set of psychological factors regarding the human behavior then instead of using only some discreptive statisical analysis e.g.( mean, standard deviation), the model can be further analyzed more intensively with the use of the structural equation modeling.
While the individual items that are used to construct measures in the HBM are ordinal, they are typically combined into multi-item scales that meet the requirement for use as interval-level variables.
So does that mean that calculating HBM constructs as mean scores and standard deviation is right? And calculating the constructs as categories eg high and low and expressed as proportions is wrong? @David L Morgan.
It is much more efficient and effective to report an overall scale score for each of the major components of the model, rather than reporting high versus low scores for each of the separate variables that go into each major component.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS