Hello. Two years ago I thought "How do I find out what high-performing students have in common?". So my friends and I collected data, but I only analyzed it now. Well... here's what the data says about sleep, self-study hours, working hours, grit, and more.
About the dataset
Characteristics and Habits
Note-taking habits
In this data set, most students prefer digital and handwritten notes i.e. on tablets with a digital pencil, followed by paper and handwritten, digital and typed, and finally paper which is typed and printed. Interestingly, both forms of digital notes have approximately the same average grade. Using paper (typed and printed) notes, e.g. book chapters, seem to have a slightly higher average grade.
Reviewing content materials after class
Almost 50% of the students review their lecture materials sometimes, which is great because it reflects an average GPA of 3.336. The highest average GPA is amongst those who always review. The lowest average GPA is amongst those who never review. Not surprising.
Self-study and Sleep
You've seen the data and the general habits by grades. Now you must be wondering whether we can predict an increase or decrease in one's grade by changing certain independent variables such as daily sleep. So, read on, friends. You're going to love what you see.
Note on correlation: there is a weak correlation between the numeric I variables of this dataset.
Regression
Context
NewProject = "New ideas and projects sometimes distract me from previous ones."
Setback = "Setbacks don’t discourage me."
LostInterest = "I have been obsessed with a certain idea or project for a short time but later lost interest."
Hardwork = "I am a hard worker."
ChangeGoal = "I often set a goal but later choose to pursue a different one."
NoFocus = "I have difficulty maintaining my focus on projects that take more than a few months to complete."
Finish = "I finish whatever I begin."
Diligent = "I am diligent"
Can the Grit Variables Predict Grades?
The distribution appears to be left-skewed (note the minimum residuals).
Statistics note: The estimate tells us the average increase in the response variable associated with a one-unit increase in the predictor variable, assuming all other predictor variables are held constant.
From this multiple regression output, we can say that the three statistically significant variables reveal that:
Sticking to your goal can increase your grades by 0.245 (on a 4.0 scale) on average.
Often leaving projects incomplete can decrease your grades by 0.184 on average.
If you're not diligent with your work, your grades can drop by an average of 0.366 grade points!!! That's almost 9%!
But... instead of trying to find the best model on my own, I used a special function in R. The `step()` function is useful for backward multiple regression testing as it automates the process of variable selection by iteratively removing or adding variables based on certain criteria like the AIC.
This helped me simplify the model selection process, ultimately saving time and reducing the chances of human bias.
Here's what I found:
As you can see, many of the variables were dropped.
This model roughly provides the same conclusion as I outlined above, except the numbers are different. Why? Think about the model fit and how the different variables within might be correlated to one another as well.
Overall, our conclusions change to:
Sticking to your goal can increase your grades by 0.23 (on a 4.0 scale) on average.
Often leaving projects incomplete can decrease your grades by 0.209 on average.
If you're not diligent with your work, your grades can drop by an average of 0.355 grade points!!!
I know that the R-squared value is rather small, so take what you want from this with a grain of salt.
What about personal habits?
Interesting output...
Study for one hour more and you can expect an average grade increase of 0.064.
Sleep for one hour more and you can expect an increase of 0.065 on average.
But take a look at 'Review After Class, No Rarely'. Weird, right? Did I find some geniuses for the data set or what? Honestly, these are business school students so I guess it makes sense... wait, this is confirmation bias!
So that's where my data skills pack their suitcases and leave. I'll leave the interpretation up to you.
Lastly, to do the step function here as well...
Yup, that makes sense.
One hour increase in self-studying can increase your grade by 0.063.
A one-hour increase in sleep can increase your grade by 0.0425.
Disclaimer
The dataset was not large enough to help me derive solid insights. There is probably a lot of confirmation bias and availability heuristic here. So take everything with a grain of salt, be data-informed, not data-driven (because the data here is not... it), use your judgment, and just have fun in life.
Anyway, I hope you enjoyed this! If you would like to request a copy of my Tableau dashboard or R script, feel free to leave a comment or email me. If you find any mistakes in my analysis or would like to offer feedback, please do reach out. I'm also on a learning journey and would appreciate your feedback to improve myself.
Bye! <3
crazy! very informative and thorough research