We want to find what questions in survey are highly correlated to overall score. We picked a group of questions based on practical meaning and calculated the average score of these questions for each survey. The correlation between the average score and overall score is pretty high. Here are the 3 questions: 1. How do I determine the cause of the high correlation? Is it because these questions really have big influence on overall score or just because the average score is the result of two many questions (it's clear that the correlation between the average score of all questions and overall score is 1)? 2. How can I find a group of questions that are really high_correlated to overall score? ( I know the correlation between each individual question and overall score) 3. Are there any good references on correlation analysis?
Please send your help to firstname.lastname@example.org. Thanks in advance.