Greetings, I would be grateful for your feedback and any references if you have them. My question is 'does a bigger sample size help the discrimination process of different (but similar) constructs during factor analysis?. I have read papers where researchers have empirically discriminated between very similar and likely highly-interrelated constructs (such as 'affective commitment to a firm' and 'affective commitment to their key account manager' ). However such studies tend to have sample sizes in the high hundreds or even thousands and I am wondering if sample size helps this.
Many thanks
You can perform HTMT discriminant analysis to determine what is also conceptually and statistically different.
Article A New Criterion for Assessing Discriminant Validity in Varia...
In general, the larger the sample size the better. Sample size is directly related to statistical power. Higher power (larger sample size) makes it more likely that you will find statistically significant results in your sample if they exist in the population from which you have sampled. Assuring that your sample is drawn from a specific population is necessary for you to draw appropriate inferences from, and interpretations of, your data.
Nick W Williams It is well known that higher the sample size more robust the results are. However, before drawing statistical inferences, validity and reliability tests can help to reduce the chances of similarity. A discriminant analysis or a CART model testing is also helpful.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size