This is my first time posting on this medium, and I’m really excited to gain constructive feedback.
I found the project quite interesting. My approach looks quite different from what I have seen, and I hope it doesn’t come off as simplistic. My estimated time of completion was 3 hours.
Please let me know what you think.
GitHub repo link: Github
Here are some comments. I will focus on corrections first:
- You said:
With Pearson correlation coefficient of 0.34, we can conclude that a moderately strong postive linear relationship exists
Actually 0.34 in some literature is considered as “weak” correlation. Source: pearsons.pdf (statstutor.ac.uk)
Correlation is an effect size and so we can verbally describe the strength of the
correlation using the guide that Evans (1996) suggests for the absolute value of r:
- .00-.19 “very weak”
- .20-.39 “weak”
- .40-.59 “moderate”
- .60-.79 “strong”
- .80-1.0 “very strong”
For example a correlation value of would be a “moderate positive
- When you classified countries as rich and poor by whether they were above and below the median GDP, you do not present which countries actually fell above and below the median. The only country classification we learn about is Chile, which is a poor country.
- Could be nice to see the strip plot and boxplot side by side
Now for the positives:
- Nice writing - it can get past a grammar nazi
- The way you describe the data manipulation is great and sounds proper
- Good quality of the visuals in terms of design and in terms of them not being blurry or unclear
Additional Analysis you could do:
- Look at the individual correlations of GDP vs Life Expectancy for the countries. For example only China. You will see that the correlatons are very strong reaching above 0.9.
Thank you very much for your feedback. My comment on the Pearson Coefficient derived was due to this article.
I recently reviewed my workbook on this project thanks to feedback from this forum. Please kindly look through and please share your thoughts.
PS: Although mentioned in the workbook, I will take every opportunity to thank Alexander Lacson and Niko (nikkibeach) for their very constructive feedback on my assumptions on this project.