Mean Correlation (Average Correlation) using Reliability Analysis in SPSS
HTML-код
- Опубликовано: 15 июл 2024
- This video demonstrates how to use the Reliability Analysis function in SPSS to produce a mean correlation (average correlation) from a correlation matrix. A comparison is made between how SPSS and Excel calculate a mean correlation.
I appreciate that you showed us how to do the correlation analysis even tho it’s didn’t specifically give us that in the first matrix. That makes it much easier to solve for.
I like how you explain things in an easy to understand way. Thanks.
Definitely doesn't seem as hard as I assumed in the beginning. The step by step instructions make it much easier to follow along with and understand.
There are so many ways to obtain results it can be mind boggling. Having a good knowledge of Excel and SPSS seems to be key but also knowing and understanding the results is even more important so that when something looks skewed, you can check yourself using the various tools available. Thanks for the Video!
I agree John. If using Excel to calculate the mean I would have not know it was incorrect. It's important to be proficient in both systems so it can be used as a way to cross reference data. I also appreciate that running one function gives a wide range of statistics. SPSS seems like it is very functional and time efficient to use when analyzing data and research.
I agree with John! Attempting to collect data can become difficult. However, this video was helpful and I now have a better undertanding of mean correlation.
Nice, thanks!
hello. thank you for all your video. I have a doubt that i can not find an answer. So i have one variable that is composed by 4 items however one of then i had to sort because the normality was sk. so my question is can I create a variable with this different item (treated and not treated) ?
correlation need to be performed on mean values or replicate values?
which one would be better?
Salutes. I have doubts with this technique. To calculate means of correlations coefficients, you can not simply use arithmetic mean formulas. Because correlation coefficient is a standardized value which can vary between -1 and +1. For example, assume that we have 3 c. coefficients (r): 0,1; -0,3; 08. According to arithmetic mean calculation, average value of these three correlations should be (1-3+8)/3= 0,2. Furthermore, let's take the absolute value of -,3 and calculate again: (1+3+8)/3 = 0,4. Instead of this, many researchers and textbooks suggest to use Fisher's transformations because of the things that I mentioned earlier this comment. When we transform these values to fishers z and calculate the average again, we will see 0,502822414. So you can see the difference. Also, I am really curious that do SPSS and Excel take the negative coefficients as absolute values.
Note: I calculated fisher values with the excel sheet of Jamie Decoster and Anne-Marie Iselin
This and other Excel spreadsheets are available at www.stat-help.com