I'm assuming you refer to the 2 negative loadings. Are these two variables reverse coded relative to (and negatively correlated with) the other variables that load on the same factor? If so, then recoding these variables and rerunning the PCA would result in positive loadings of the same magnitude.
@Samanta Negative factor loadings are generally indication of negative polarity items which have not been reverse coded before running EFA. Sometimes, it's because of biased response as well.
I agree with the previously provided observations. I suggest provision of the full loading matrix in the future because the additional information available in that matrix would probably have been informative for addressing your question. I suppose the suppression everything except primary loadings provides clarity on some accounts but not so much on other accounts. My impression is, however, that you've received sound guidance in any event.
- Badges
- Science topic