Analysis and Conclusion of Student P5(2016-2017) with SAS

This article conclude the study with P5(2016-2017) & P6(2017-2018) base on the previous analysis.

Foreword

1. The target of this report is the students of 2016-2017 P5 & 2017-2018 P6.
2. There is no description of the data process and source code of SAS. If you want to check, please click this link.
3. Currently, there are just some simple report and diagram base on the part of statistical indicators, furthermore, some conclusion together the teacher’s opinions were attached.
4. Later, if time is enough, I will add more comprehensive statistics.

1. Base on Different Classes in Each Grade

There are two classes in 2016-2017 P5, which contain around 50 students.

After they upgraded to P6, 50 students were mixed randomly and divided into two classes in 2017-2018 P6.

1.1 Class Comparision in P5 base on Mean

Figure: Class Comparision in P5 base on Mean

Conclusion:

1. The mean of Class P5-2 score is distinctly higher than that of Class P5-1.
2. However, the stand deviation (which quantify the amount of variation of a set of data values) of P5-2 is higher than that of P5-1 as well. Namely, the gap between good scores and poor scores in P5-2 is more prominent.
3. the increment of STD of P5-2 is gradually higher for each term. Namely, in each term, the gap of scores in P5-2 was growing.
4. the STD of term1 in P5-1 is still not quite big; however, since from term2, the STD is growing up obviously (the pace of increment is lower than that of P5-2).
5. in summary, the means score P5-1 is tiny lower than P5-2, but more stable, and uniform.

1.2 Class Comparision in P6 base on Mean

Figure: Class Comparision in P6 base on Mean

Conclusion:

1. Except Term 2 in Grade 6, the mean of Class P6-1 score is a little bit of higher than that of Class P6-2.
2. In summary, the situation is happened to reverse with that in the last academic year(when P5)

2. Base on All Students performance in Each Grade

Different from the above one, the statistics in this chapter will be base on all students performance in each grade. It will analyze the level and some statistical indicators base on all student score.

2.1 Overall Stat of P5 & P6 with box

Figure: Overall Stat of P5 & P6 with box

Conclusion:

1. During P5, the deviation was more and more obvious followed the development of terms.
2. The bias of mean of all terms in P5 and P6 is not significant.
3. The performance of P6 is worse than that P5.

2.2 P5 Score Distribution

Figure: P5 Score Distribution

Conclusion:

Followed the term forward, the excellent score(above 92.5) became more and more. Meanwhile, the percentage of the lower score(lower than 72.5) almost didn’t change. Namely, the variation became bigger.

2.3 P6 Score Distribution

Figure: P6 Score Distribution

Conclusion:

Variation is smaller compared with that in P5. Meanwhile, as it was displayed in sub-chapter 2.1, we can know that the overall of P6 score is worse than P5.
(It seems like the students in P6 don’t like to study any more….)

2.3 P5 & P6 Score Distribution

Figure: P5 & P6 Score Distribution

Conclusion:

Good became beeter, bad became worse.

3. Base on each Student in Two Academic Years

3.1 Most unstable score

Figure: Most unstable score

Conclusion:
There are some students whose score of 8 terms in P5 & P6 fluctuates with a distinct range. However, their mean score is not high, even lower. That means student sometimes got a high score, but sometimes score plummets, and finally, the mean score is not so high.

3.2 Most stable score

Figure: Most stable score

Conclusion:

good ones are always good; worse ones remain worse.

4. Base on Trend-featured Student in Two Academic Years

4.1 Trend of going up

In the previous article, I already introduced the method and got the thumbnail of figures like this:

Figure: Trend of going up

then we can take the obvious ones to analyze using the following code:

PROC SGPANEL DATA=ss_16_17.tempT4series_goup(where=(name in ("GEOFFREY MIKHAEL","JEREMY JERO","JAMES DYLAN IRWANTO"))) noautolegend;
	PANELBY trend name  / NOVARNAME COLUMNS= 2 ONEPANEL;
series x=term y=score1 / GROUP=name datalabel;
COLAXIS grid; 
run;

Figure: take the obvious ones

Conclusion:

• from the above figure, we can see JAMES among these three students is the one who improved significantly and became stable.
• both of GEOFFREY and JEREMY have significant improvement, however, GEOFFREY had the steady growth in P5 and his score fluctuated in P6. JEREMY had the opposite situation.

4.2 Trend of going down

After I chose the three typical featured students, the figures are shown as below:

Figure: Trend of going down

Conclusion:

• NELSON and REINALDO has a similar situation in P6 phase: after enter P6, the score plummeted. This feature was shown on REINALDO more obviously, as his P5 score was in the range of middle above.
• LIONEL’s score was going down from the first term in P5 and almost touched the lowest score - 70. in P6, seems like he wasn’t will to study any more, namely giving up.

4.3 Scores fluctuated greatly

In this part, I chose nine students whose score fluctuated significantly in the period of P5 and P6.

Figure: Scores fluctuated greatly

I concluded these following situation from the above figures:

1. Ones had significant improvement in P5 but plummeted in P6, such as KENZIE, KIMBERLY and AILEEN.
2. Ones had mediocre scores in P5, then raise in P6 and plummeted soon, such as WINSTON, JENSON.
3. the score likes a roller coaster, up and down alternatively, such as AILEEN and HADRIAN.