How to improve your child high school exam performance?
A data-based approach using ‘Students Performance in Exams’ dataset
As parents, we always want the best for our children. Performing well at school is one of the major factors helping our children to achieve a brighter future. It can be said that exam performance of one’s child is a good predictor of their future. If not, at least doing well in exam makes us parents feel proud. Now, I have a personal viewpoint on the reality of these statements. It’s likely that you may as well have your own opinion. But what does the data suggest?
By courtesy of the ‘Students Performance in Exams’ dataset which is available on Kagglet, this will help us in gaining insight on how the student’s performance (test scores) is affected by a variety of personal, social, and economic factors (Gender, Ethnicity, Parental level of education, Lunch type, Test preparation course). One among the interesting questions that intrigued me was whether one’s parent level of education had any impact at all on their offspring’s exam performance.
Dataset
The dataset consists of math, reading and writing scores of students along with some influencing factors such as gender, race, parental level of education, lunch type, and if the student took test preparation courses. The data are from 1000 high school students from the United States
Here, you can see all the column names, total values and type of the values.
We have 2 types of variables.
- Numerical variables: which contain numbers as values
- Categorical variables: which contain descriptions of groups or things.
In this Data set,
Numerical Variables are Math score, Reading score and Writing score.
Categorical Variables are Gender, Race/ethnicity, Parental level of education, Lunch and Test preparation course.
So, does gender matter in high school exam performance?
Does it matter if it’s a girl or a boy, do either of them performs better at exam?
Firstly, the data almost evenly balanced between male and female as it can be seen in the diagram below
From the visualization and table above, we find that females have better reading and writing scores which leads to better total score than males. Males on the contrary to females have inferior total score but a better math score.
Could it be due to the fact that math is more intuitive, more practical and requires less of study? Maybe! But we cannot say for sure. All we can infer from the data is females have higher total scores, but males have higher math scores.
What about ethnicity? Does it matter?
In my opinions, ethnicity had nothing to do with high school exam performance. But what does the data say?
The students belong to different ethnicity groups, ranging from A to E. These represent different ethnicity group which are unknown.
But it could be clearly seen that race/ethnicity group E had better scores in all the subjects leading to a better total score. The race/ethnicity group A had the worst exam performance. It can be noted that most groups performed relatively poor in math compared to reading and writing with the exception of group A which had it’s best score in math compared to reading and writing. Three groups namely A, B and C had their best score in reading.
Does completing a test preparation course leads to better performance?
As you can see from the plot above, completing a test preparation course led to a better total score. I think this is intuitive as the more you prepare and practice the better are the exam performances.
Does Lunch type have any effect on exam performance?
In the US, school meal programs provide school meals free of charge, or at a government-subsidized price, to U.S. students from low-income families. You can read further on this and on National School Lunch Act on the link below.
So does the lunch type matter?
This shows that students that had standard lunch perform better than those provided with free/reduced lunch at school. It could also be interpreted that students having standard lunch are more from middle-class or higher income families and as such they might afford better study materials. Another argument could be that the two types of lunch had different nutritional values which could in term influence the academical performance showing that diet may play a role in performance.
Does parental level of education affect their offsprings performance?
Well, this is an intriguing question! Does one having a master degree or any kind of education even matter when it comes to their child’s exam performance?
According to the data, it does matters. The data shows that those students whose parents had a higher degree also performed better at exam.
Students whose parents had a master’s degree had the highest score in all categories, followed by those having a bachelor’s degree and an associate degree. It could be noted that on overall, children whose parents studied only till high school performed the least at exam.
One plausible argument could be that students whose parents had a higher education might have had a higher income, thus providing them with an improved environment including test preparation courses and also better diet (Lunch) leading to better performance.
Conclusion
Obviously, there are factors like race/ethnicity and gender that are beyond the human capabilities to influence. But according to this data analysis, we spotted other factors that could potentially impact our children academical performance such as level of one’s education, test preparation course and type of lunch.
As such if you have the opportunity to pursue further education, then go for it as it may not only enhance your career but will also positively influence your child’s exam performance, the higher your level of education, the better your child’s exam performance.
Also, the type of facilities we provide to our children seem to influence their exam performance, thus if you can afford test preparation course and standard lunch instead of free/reduced one, then you should probably opt to do so.
