Evaluation Methods: Why Randomize?

Submitted by jyuan@worldbank.org on Tue, 04/12/2016 - 18:12
Image
evaluationmethods

Course Overview

1.What is evaluation?

2.Measuring impacts (outcomes, indicators)

3.Why randomize?

4.How to randomize?

5.Sampling and sample size

6.Threats and Analysis

7.Cost-Effectiveness Analysis

8.Project from Start to Finish

 

Evaluation Methods from clearsateam

Download File
Pdf as plain
Evaluation Methods
1. Why Randomize?
2. Course Overview 1. What is evaluation? 2. Measuring impacts (outcomes, indicators) 3. Why randomize? 4. How to randomize? 5. Sampling and sample size 6. Threats and Analysis 7. Cost-Effectiveness Analysis 8. Project from Start to Finish
3. What is the most convincing argument you have heard against RCTs? A. Too expensive B. Takes too long C. Unethical D. Too difficult to design/implement E. Not externally valid (Not generalizable) F. Can tell us whether there is impact, and the magnitude of that impact, but not why or how (it is a black box) 0% 0% 0% 0% 0% Too expensive Takes too long Unethical Too difficult to design/... Not externally valid (No...
4. Impact: What is it? A. Positive B. Negative C. No impact D. Don’t Know 0% 0% 0% 0% Positive Negative No impact Don’t Know Intervention Primary Outcome Time
5. Impact: What is it? Time Primary Outcome Impact Intervention
6. Impact: What is it? A. Positive B. Negative C. No impact D. Don’t Know 0% 0% 0% 0% Positive Negative No impact Don’t Know Primary Outcome Intervention Time
7. Impact: What is it? Time Primary Outcome Intervention Impact
8. Impact: What is it? Time Primary Outcome Impact Intervention
9. How to Measure Impact?  Impact is defined as a comparison between:  The outcome some time after the program has been introduced  The outcome at that same point in time had the program not been introduced  This is know as the “Counterfactual”
10. Counterfactual  The Counterfactual represents the state of the world that program participants would have experienced in the absence of the program (i.e. had they not participated in the program)  Problem: Counterfactual cannot be observed  Solution: We need to “mimic” or construct the counterfactual
11. IMPACT EVALUATION METHODS
12. Impact Evaluation Methods 1. Randomized Experiments Also known as:  Random Assignment Studies  Randomized Field Trials  Social Experiments  Randomized Controlled Trials (RCTs)  Randomized Controlled Experiments
13. Impact Evaluation Methods 2. Non- or Quasi-Experimental Methods  Pre-Post  Simple Difference  Differences-in-Differences  Multivariate Regression  Statistical Matching  Interrupted Time Series  Instrumental Variables  Regression Discontinuity
14. WHAT IS A RANDOMIZED EXPERIMENT?
15. The Basics  Start with simple case:  Take a sample of program applicants • Randomly assign them to either: • Treatment Group – is offered treatment • Control Group - not allowed to receive treatment (during the evaluation period)
16. Key Advantage  Because members of the groups (treatment and control) do not differ systematically at the outset of the experiment,  Any difference that subsequently arises between them can be attributed to the program rather than to other factors.
17. WHY RANDOMIZE?
18. Example: Pratham’s Balsakhi Program Case 2: Remedial Education in India Evaluating the Balsakhi Program
19. What was the Problem?  Many children in 3rd and 4th standard were not even at the 1st standard level of competency  Class sizes were large  Social distance between teacher and many of the students was large
20. Context and Partner  124 Municipal Schools in Vadodara (Western India)  2002 & 2003:Two academic years  ~ 17,000 children  “Every child in school and learning well”  Works with most states in India reaching millions of children
21. Proposed Solution  Hire local women (Balsakhis)  From the community  Train them to teach remedial competencies • Basic literacy, numeracy  Identify lowest performing 3rd and 4th standard students • Take these students out of class (2 hours/day) • Balsakhi teaches them basic competencies
22. Pros  Reduced social distance  Reduced class size  Teaching at appropriate level  Improved learning for lower-performing students  Improved learning for higher-performers Cons  Less qualified  Teacher resentment  Reduced interaction with higher-performing peers  Increased gap in learning  Reduced test scores for all kids Possible Outcomes What is the Impact?
23. J-PAL Conducts a Test at the End  Balsakhi students score an average of 51% What can we conclude?
24. 1. Pre-post (Before vs. After) Average change in the outcome of interest before and after the programme  Look at average change in test scores over the school year for the balsakhi children
25. Method 1: Pre vs Post (Before vs. After) Average test scores of Balsakhi 24.8 51.22 60 50 40 30 20 10 0 students 26.42 Start of program End of program Average post-test score for children with a Balsakhi 51.22 Average pretest score for children with a Balsakhi 24.80 Difference 26.42
26. Pre-Post  Limitations of the method • No comparison group, doesn’t take time trend into account What else can we do to estimate impact?
27. Method 2: Simple Difference Measure difference between program participants and non-participants after the program is completed Divide the population into two groups: One group enrolled in Balsakhi program (Treatment) One group not enrolled in Balsakhi program (Control) Compare test score of these two groups at the end of the program.
28. Method 2: Simple Difference Average test scores end of program 56.27 51.22 60 50 40 30 20 10 0 -5.05 Not enrolled in program Enrolled in program Average score for children with a balsakhi 51.22 Average score for children without a balsakhi 56.27 Difference -5.05 QUESTION: Under what conditions can the difference of -5.05 be interpreted as the impact of the Balsakhi program?
29. Method 3: Difference-in-difference Measure improvement (change) over time of participants relative to the improvement (change) over time of non-participants  Divide the population into two groups: • One group enrolled in Balsakhi program (Treatment) • One group not enrolled in Balsakhi program (Control)  Compare the change in test scores between Treatment and Control • i.e., difference in differences in test scores  Same thing: compare difference in test scores at post-test with difference in test scores at pretest
30. Method 3: Difference-in-difference 24.8 Average test scores 51.22 36.67 56.27 60 50 40 30 20 10 0 Start of program End of program Enrolled in Balsakhi program Not enrolled in Balsahki program Pretest Post-test Difference Average score for children with a Balsakhi 24.80 51.22 26.42
31. Method 3: What would have Happened without Balsakhi? Method 3: Difference-in-differences 26.42 75 50 25 0 2002 2003
32. Method 3: Difference-in-differences Pretest Post-test Difference Average score for children with a balsakhi 24.80 51.22 26.42 Average score for children without a Balsakhi 36.67 56.27 19.60
33. Method 3: What would have Happened without Balsakhi? Method 3: Difference-in-differences 26.42 19.60 6.82 points? 75 50 25 00 2002 2003
34. Method 3: Difference-in-Differences  QUESTION: Under what conditions can 6.82 be interpreted as the impact of the balsakhi program?  Issues: • failure of “parallel trend assumption”, i.e. impact of time on both groups is not similar Pretest Post-test Difference Average score for children with a Balsakhi 24.80 51.22 26.42 Average score for children without a Balsakhi 36.67 56.27 19.60 Difference 6.82
35. Method 4: Regression Analysis  Divide the population into two groups: • One group enrolled in Balsakhi program • One group not enrolled in Balsakhi program  Compare test score of these two groups at the start and at the end of the program.  Control for additional variables like gender, class-size  Post-test = 훽0 + 훽1Pre−test + 훽2Gender + 훽3Class−size + 훽4퐵푎푙푠푎푘ℎ푖 + 푒
36. Method 4: Regression Analysis Income 0 20 40 60 80 Test Score (at Post Test) 1.92 post_tot_noB post_tot_B Linear (post_tot_noB) Linear (post_tot_B) QUESTION: Under what conditions can the coefficient of 1.92 be interpreted as the impact of the Balsakhi program?
37. Impact of Balsakhi Program 26.42 -5.05 6.82 1.92 30 25 20 15 10 5 0 -5 -10 (1) Pre-post * (2) Simple Difference * (3) Difference-in-Difference * (4) Regression with controls * Significant at 5% level Method Impact Estimate (1) Pre-post 26.42* (2) Simple Difference -5.05* (3) Difference-in-Difference 6.82* (4) Regression with controls 1.92
38.  Counterfactual is often constructed by selecting a group not affected by the program  Non-randomized: • Argue that a certain excluded group mimics the counterfactual.  Randomized: • Use random assignment of the program to create a control group which mimics the counterfactual. 38 Constructing the Counterfactual
39. Randomised Evaluations Individuals, villages, or districts are randomly selected to receive the treatment, while other villages serve as a comparison Groups are Statistically Identical before the Program Treatment Group Comparison Group Village 2 Village 1 = Any Difference at the Endline can be Attributed to the Program Two groups continue to be identical, except for treatment. Later, compare outcomes (health, test scores) between the two groups. Any differences between the groups can be attributed to the program.
40. Basic Set-up of a Randomized Evaluation Target Population Not in evaluation Evaluation Sample Total Population Random Assignment Treatment Group Control Group
41. Random Sampling and Random Assignment Randomly sample from area of interest
42. Random Sampling and Random Assignment Randomly sample from area of interest Randomly assign to treatment and control Randomly sample from both treatment and control
43. Randomization Design  Population = all schools in case villages  Target population: weakest students in all of these schools  Stratify on three criteria: • Pre-test scores • Gender • Language  Give 50% of them the Balsakhi program
44. Impact of Balsakhi - Summary Method Impact Estimate (1) Pre-post 26.42* (2) Simple Difference -5.05* (3) Difference-in-Difference 6.82* (4) Regression 1.92 *: Statistically significant at the 5% level
45. Which of these methods do you think is closest to the truth? A. Pre-post B. Simple difference C. Difference-in-Difference D. Regression E. Don’t know Method Impact Estimate (1) Pre-post 26.42* (2) Simple Difference -5.05* (3) Difference-in-Difference 6.82* (4) Regression 1.92 *: Statistically significant at the 5% level
46. Impact of Balsakhi - Summary Method Impact Estimate (1) Pre-post 26.42* (2) Simple Difference -5.05* (3) Difference-in-Difference 6.82* (4) Regression 1.92 (5)Randomized Experiment 5.87* *: Statistically significant at the 5% level
47. Example #2 - Pratham’s Read India Program Method Impact (1) Pre-Post 0.60* (2) Simple Difference -0.90* (3) Difference-in-Differences 0.31* (4) Regression 0.06 *: Statistically significant at the 5% level
48. Which of these methods do you think is closest to the truth? Method Impact (1) Pre-Post 0.60* (2) Simple Difference -0.90* (3) Difference-in- Differences 0.31* (4) Regression 0.06 *: Statistically significant at the 5% level A. Pre-post B. Simple difference C. Difference-in-Difference D. Regression E. Don’t know 0% 0% 0% 0% 0% A. B. C. D. E.
49. Example #2 – Pratham’s Read India Program Method Impact (1) Pre-Post 0.60* (2) Simple Difference -0.90* (3) Difference-in-Differences 0.31* (4) Regression 0.06 (5) Randomized Experiment 0.88* *: Statistically significant at the 5% level
50. Method Comparison Works only if… Pre-Post Program participants before program Nothing else was affecting outcome Simple Difference Individuals who did not participate (data collected after program) Non-participants are exactly equal to participants Differences-in- Difference Same as above + data collected before and after If two groups have exactly the same trajectory over time Regression Same as above + additional “explanatory” variables Omitted variables do not affect results Randomized Evaluation Participants randomly assigned to control group The two groups are statistically identical on observed and unobserved characteristics Summary of Methods
51. Conditions Required Method Comparison Group Works if…. Pre-Post Program participants before program The program was the only factor influencing any changes in the measured outcome over time Simple Difference Individuals who did not participate (data collected after program) Non-participants are identical to participants except for program participation, and were equally likely to enter program before it started. Differences in Differences Same as above, plus: data collected before and after If the program didn’t exist, the two groups would have had identical trajectories over this period. Multivariate Regression Same as above plus: Also have additional “explanatory” variables Omitted (because not measured or not observed) variables do not bias the results because they are either: uncorrelated with the outcome, or do not differ between participants and non-participants Propensity Score Matching Non-participants who have mix of characteristics which predict that they would be as likely to participate as participants Same as above Randomized Evaluation Participants randomly assigned to control group Randomization “works” – the two groups are statistically identical on observed and unobserved characteristics
52. Other Methods  There are more sophisticated non-experimental methods to estimate program impacts: • Regression • Matching • Instrumental Variables • Regression Discontinuity  These methods rely on being able to “mimic” the counterfactual under certain assumptions  Problem: Assumptions are not testable
53. Conclusions: Why Randomize?  There are many ways to estimate a program’s impact  This course argues in favor of one: randomized experiments • Conceptual argument: If properly designed and conducted, randomized experiments provide the most credible method to estimate the impact of a program • Empirical argument: Different methods can generate different impact estimates
54. Key Steps in Conducting an Experiment 1. Design the study carefully 2. Randomly assign people to treatment or control 3. Collect baseline data 4. Verify that assignment looks random 5. Monitor process so that integrity of experiment is not compromised 6. Collect follow-up data for both the treatment and control groups 7. Estimate program impacts by comparing mean outcomes of treatment group vs. mean outcomes of control group. 8. Assess whether program impacts are statistically significant and practically significant.
55. THANK YOU
Date
Regional Center Tag
Resource Type Tag
Language Type Tag
Image Thumb
S