Breakout Session Task 1

Instructions:
In this breakout session, we will go over the impact of a TVET programme on youth employment in Uganda. Begin by reading the provided information about the
TVET programme and the research questions guiding the impact evaluation. As a group discuss each question using the information provided and the tables to
inform your discussion. Select one person from your group who will present the results. A moderator has been assigned to each group and will help facilitate the
discussions. Use the provided answer sheet template to record your groups’ answers to each question. Any additional insights or comments can also be written
in the answer sheet.

Theme: Descriptive
The table below presents key data for several variables we are interested in for a counterfactual impact evaluation on a TVET programme in Uganda. The trainings
were conducted for youth between the ages of 19-30 years within the West Nile Region in Uganda. The general objective of the programme was to provide the
youth with market relevant skills to improve their chances of finding employment and increasing their income. The impact evaluation is guided by the following
evaluation questions:

1. Evaluation Question 1: Did the TVET programme have an impact on the employment status of trainees?
2. Evaluation Question 2: Did the TVET programme have an impact on the income of trainees?
3. Evaluation Question 3: Did the TVET programme have an impact on the self-perceived employability of trainees?

The data was captured 18 months after participants completed the technical skills training. This data includes both the intervention group (treatment group)
made of participants in the TVET programme and non-intervention group (control group) which comprises applicants who applied but were not selected for the
training and, as a result, did not undergo the technical skills training. Based on the research questions, our primary focus lies on several variables of interest
among respondents including gender, age, educational level, employment status, income, and self-perceived employability.
Table 1
Employment status

Self Employee Apprenticeship Unemployed Total

Employed/Employer
N % N % N % N % N %
Gender Male 51 6.7 427 56.3 61 8.0 219 28.9 758 100
Female 97 16.3 295 49.6 94 15.8 109 18.3 595 100
Refugees Refugee 10 3.0 112 33.7 9 2.7 201 60.5 332 100
Non-Refugee 138 13.5 610 59.7 146 14.3 127 12.4 1021 100
Educational status Primary 43 12.9 101 30.2 84 25.1 106 31.7 334 100
Secondary 64 9.6 380 57.0 59 8.8 164 24.6 667 100
Tertiary 41 12.0 241 70.3 3 0.9 58 16.9 343 100

Table 2
Variable N Mean SD Min Max Median Mode T-test mean p-value
difference (T-C)
Employment status 1353 0.65 0 1 0.6 0.2 0.01**
Income (Ugandan Schilling) 1353 475 76.8 300 700 480 450 45.6 0.021*
Self-perceived employability (1-5) 1353 3.9 0.8 2 5 4 4 0.1 0.456
Age (years) 1353 24.5 4.5 18 35 25 24 1.1 0.643
Significance stars: * p < 0.05, ** p < 0.01

1. Based on Table 1, what insights can we draw from the distribution of employment status among the different subcategories in our sample?
2. Based on Table 2, can you identify potential outliers in any of the variables and explain why they might be outliers?
3. How might these outliers impact our overall analysis and how can these be handled? What should we be cautious about when dealing with them?
4. Besides the outliers are there any other factors we should take into account to understand better the data we have?
5. Before we look further into the data we would like to understand the data better by looking at the difference in the means of our sample. But before we
do that we would need to come up with our null hypothesis:
a. Hypothesis 1 (Employment)
i. H0 :
 b. Hypothesis 2 (Income)
i. H0 :
c. Hypothesis 3 (Self-perceived employability)
i. H0 :
6. What do the T-tests and p value associated with the T-test mean for employment status, income and self-perceived employability?

7. Based on the following Table 1 and Table 2, can we conclude that the TVET programme has impacted on the following indicators
a. Employment status
b. Income
c. Self-perceived employability
8. Based on the table of results are we able to explain if the TVET programme have had an impact on income and self-perceived employability?
 Breakout Session Task 2:
Instructions:
In this breakout session, we will go over the impact of a TVET programme on youth employment in Uganda. Begin by reading the provided information about the
TVET programme and the research questions guiding the impact evaluation. As a group discuss each question using the information provided and the tables to
inform your discussion. Select one person from your group who will present the results. A moderator has been assigned to each group and will help facilitate the
discussions. Use the provided answer sheet template to record your groups’ answers to each question. Any additional insights or comments can also be written
in the answer sheet.

Theme: Statistical testing

The table below presents key data for several variables we are interested in for a counterfactual impact evaluation on a TVET programme in Uganda. The trainings
were conducted for youth between the ages of 19-30 years within the West Nile Region in Uganda. The general objective of the programme was to provide the
youth with market relevant skills to improve their chances of finding employment and increasing their income. The impact evaluation is guided by the following
evaluation questions:

1. Evaluation Question 1: Did the TVET programme have an impact on the employment status of trainees?
2. Evaluation Question 2: Did the TVET programme have an impact on the income of trainees?
3. Evaluation Question 3: Did the TVET programme have an impact on the self-perceived employability of trainees?

The data was captured 18 months after participants completed the technical skills training. This data includes both the intervention group (treatment group)
made of participants in the TVET programme and non-intervention group (control group) which comprises applicants who applied but were not selected for the
training and, as a result, did not undergo the technical skills training. Based on the research questions, our primary focus lies on several variables of interest
among respondents including gender, age, educational level, employment status, income, and self-perceived employability.

In Table 3, the results of a regression analysis are displayed:

Table 3
Variable Number of Number Trainees Non-trainees regression p-value
trainees of non- mean mean coefficient (tvet)
trainees
Gender (Male) 676 216 1.0 1.0 0.03 1
Age (Years) 676 216 45.3 45.4 0.00 0.64
 Household size 676 216 5.9 5.8 -0.07 0.32
Employment status (Employed) 483 387 0.7 0.6 0.09 0.12
Income 483 387 495.7 450.1 45.67 0.02*
Self-perceived employability (1- 676 216 3.9 3.8 0.23 0.45
5)
Significance stars: * p < 0.05, ** p < 0.01

1. Observing the regression coefficients of the variables and based on our hypothesis, what do we say about the relationship between the TVET programme
and youth employment?
a. Employment status
b. Income
c. Self-perceived employability
2. At what confidence level can we reject or retain our hypothesis?
3. How do these new observations differ from the previous observations from the descriptive statistics on the TVET programme and youth employment?
4. Suggest the policy implications of the impacts of TVET programmes on youth employment based on the results above
5. Based on the previous observations and findings, how can we use qualitative methods to better understand the ways in which TVET impacted the
employment status of the beneficiaries?
- Which questions would you ask and to whom?
- How would you use the answers to strengthen the findings of the CIE?