Hello,
I replicated a study in which participants are asked to rate the importance of some user experience dimensions (like efficiency, usefulness, perspicuity, etc.) for a product (0-7 rating).
I added some new dimensions like ease of use. How can I statistically determine whether the added dimension is measuring a new dimension? It is obvious that the new dimension have strong correlation with other pragmatic dimensions. My question is how to show the new dimension is worth being included and is different from the current ones.
P.S. There is no list of items per dimension. Definition of dimensions are provided and participants just rate their importance.
Thank you
If you are interested in reading the following link, pay special attention to the meaning of the columns of the Item-Total Statistics table. You will find your answer in it.
https://graphpad.ir/cronbach-alpha-reliability/
Abolfazl Ghoodjani Thank you for your response. Based on the link you shared with me, "Cronbach's alpha if item deleted" determines an item to be removed if it makes the overall Cronbach's alpha higher. Hence, when no dimension caused an increase of Cronbach's alpha, they should all be retained. Does it also prove that none of them is similar?
آفرین
منظور من هم دقیقا همان ستون
Cronbach's alpha if item deleted
بود. البته من به یافتههای ستون
Squared Multiple Correlation
بیشتر علاقه دارم.
Thank you for your quick response Abolfazl Ghoodjani. To my understanding, the interpretation of the results of the variance inflation factor (when considering one independent variable as a dependent) should be similar to "squared multiple correlations" since VIF uses R2 for calculation. Thus, higher R2 results in multicollinearity and we need to remove one item.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics