Skip to content

KwokHing/Pencils-of-Promise

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

34 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Data for A Cause - Pencils of Promise

Goal: Visualising effectiveness of instruction in Ghana [data pre-processing work are not documented here yet]

1. Effectiveness of Teacher Support programming by analyzing changes within individual teachers' scores over time

png

The score differences (ranging from decrease of -2 to increase of 2 points) of each question (recorded from Round 1 and Round 2) for each teacher are plotted above. Most of the teacher's scoring for the different questions remained unchanged over time while some of the teacher's score increase or decrease over time. The table below details the score changes for each question over time.

Qns Num. of -2pts Decrease Num. of -1pt Decrease Num. of no diff Num. of 1pt Increase Num. of 2pts Increase
1 1 17 57 15 2
2 NA 17 55 19 1
3 5 23 43 13 7
4 2 12 62 15 1
5 1 15 56 18 2
6 1 21 54 16 NA
7 2 20 45 20 5
8 1 18 44 28 1
9 2 20 42 24 4
10 1 20 47 22 2
11 NA 13 49 29 1

From the above table and previous plot, we are not sure if there are any improvement of the teacher's scoring over time. Therefore, the mean score of each question is plotted to visualise if there are any improvement over time.

png

The histogram plot above shows that other than 'Question 6', which dips slightly, the average scores (dotted line) for all the other questions increases over time signaling some form of success of the Teachers Support progamming.

2. Analyzing average changes and average scores between Treatment vs Control group

Control Treatment
18 166

It appears that there are much more teachers in the 'Treatment' group as compared to teachers in the 'Control' group (approx. 10%)

2.1: Control vs Treatment for observation round 1

The histogram plot below shows the average scores (dotted line) and score frequencies for teachers in observation round 1. The 'Treatment' group performs pretty much better than the 'Control' group except for 'Question 8' where there are no significant difference (average difference of only 0.025, shown in the Table below) between the 'Treatment' and 'Control' group.

png

qnsRating round1ControlMean round1TreatmentMean round1MeanDiff
Q1Rating 2.4444444444444446 2.6867469879518073 0.2423025435073627
Q2Rating 2.3333333333333335 2.6144578313253013 0.2811244979919678
Q3Rating 1.7777777777777777 2.5 0.7222222222222223
Q4Rating 2.5555555555555554 2.7349397590361444 0.17938420348058903
Q5Rating 2.3333333333333335 2.7228915662650603 0.38955823293172687
Q6Rating 2.3333333333333335 2.819277108433735 0.48594377510040143
Q7Rating 2.5555555555555554 2.710843373493976 0.1552878179384205
Q8Rating 2.4444444444444446 2.4698795180722892 0.025435073627844584
Q9Rating 2.3333333333333335 2.63855421686747 0.30522088353413634
Q10Rating 2.2222222222222223 2.566265060240964 0.34404283801874147
Q11Rating 2.111111111111111 2.6144578313253013 0.5033467202141901

2.2: Control vs Treatment for observation round 2

In observation round 2, the 'Treatment' group also performs much better than the 'Control' group across all the 11 questions further signaling the effectiveness of the Teachers Support progamming.

png

qnsRating round2ControlMean round2TreatmentMean round2MeanDiff
Q1Rating 2.5555555555555554 2.7710843373493974 0.21552878179384205
Q2Rating 2.5555555555555554 2.855421686746988 0.29986613119143257
Q3Rating 2 2.6265060240963853 0.6265060240963853
Q4Rating 2.5555555555555554 2.8433734939759034 0.2878179384203481
Q5Rating 2.4444444444444446 2.8433734939759034 0.3989290495314588
Q6Rating 2.4444444444444446 2.746987951807229 0.30254350736278424
Q7Rating 2.5555555555555554 3.0240963855421685 0.46854082998661317
Q8Rating 2.6666666666666665 2.8072289156626504 0.1405622489959839
Q9Rating 2.6666666666666665 2.963855421686747 0.29718875502008046
Q10Rating 2.6666666666666665 2.8072289156626504 0.1405622489959839
Q11Rating 2.3333333333333335 2.783132530120482 0.4497991967871484