Hello Talia,

Thank you for sharing your question with me.

Here are a few of my thoughts to add to the discussion ...

In factor analysis, the choice of a threshold for item loadings is somewhat subjective and should be informed by the research context, statistical literature, and the level of precision required. Common practice suggests that loadings of 0.30 are considered to be the minimum, while loadings of 0.40 to 0.50 and above are deemed more acceptable and indicate more substantial contributions of the items to the factor. The threshold you mentioned, 0.60, suggests a more conservative approach, ensuring that each item has a strong loading on the factor, which can contribute to a more reliable and valid factor structure. This can be particularly important when constructing a new scale, as it helps to ensure the items are well-represented by the underlying construct. It is advisable to justify the chosen threshold with references to the empirical literature. For example, Hair et al. (2010) in "Multivariate Data Analysis" suggest that loadings of 0.50 or above are considered very significant. Similarly, Tabachnick & Fidell (2007) in "Using Multivariate Statistics" provide guidelines on the interpretation of factor loadings. Ultimately, the decision should also be based on the overall model fit and interpretability of the factors. If the model retains theoretical consistency and the factors are interpretable with a lower threshold, it could be justified. If the higher threshold maintains the integrity of the construct and leads to a more parsimonious model, then it could be preferable. It is also essential to consider the trade-off between a higher threshold and the potential loss of items, which could affect the breadth of the construct being measured. 

I hope this helps :) Warmest wishes, Samantha.