Hi,
I am carrying out a very basic undergraduate research study with no stats teaching on the course (medical degree) as a result I am floundering and wonder if anyone could help.
I am using two pre-validated Likert scales looking at the relationship of Emotional Intelligence and Resilience I have put this into a scatterplot in Excel to which it looks like a positively significant result (graph attached) My Pearson's correlation coefficient between the two is coming out at 0.797 would that be r=0.79? and from what I read this is significantly significant?
However when I attempt to work out the P-value it shows as 3.419 which is not.
Any assistance or thoughts on what I am doing wrong would be of great help.
Many Thanks :)
A Pearson correlation coefficient of r = .797 would be rounded to .80. But yes, it looks like the scatterplot could correspond to such a strong positive linear association (r = 1 would indicate a perfect positive linear association).
The p value (3.419) can't be correct (p values > 1 don't make sense). Could this be the value of the test statistic (t test for correlations, testing the null hypothesis that the correlation is zero in the population)? In any case, a strong correlation of .797 should be statistically significant (p < .05), even for the relatively small number of data points (small sample size) shown in your scatterplot.
Adding to Christian Geiser's comments, using Excel for serious statistical analysis is generally a bad idea. This page outlines some of the main problems:
- https://people.umass.edu/evagold/excel.html
You probably have access to one or more recognized stats packages via your institution. If not, you could use JASP or Jamovi, both of which are free and user-friendly.
- https://jasp-stats.org/
- https://www.jamovi.org/
HTH.
Thank you, I had tried to use JASP as we don't have any SPSS packages available through the institution but was advised it was not needed at this level. It would be good to include this in my limitations section though as to why my results analysis was limited, every cloud!
If you just need the p-value, you could get an online calculation:
https://www.socscistatistics.com/pvalues/pearsondistribution.aspx
HTH, Karin
Will suggest you work out /run the analysis for the p-value again. And i was wondering if you could breifly describe the process employed.
The test results indicate that there is a correlation of 0.79, Pearson's correlation with an r-value of 0.3-0.5 weakly correlated, 0.5-0.7 moderately correlated, and 0.7-0.9 strongly correlated.
See also my review here: Technical Report Farewell SPSS: Switching to free software in 2024
- Badges
- Science topic