Hello,
I want to evaluate impact of an conditional cash transfer scheme on farm income. Here, farmers who have land were eligible and all the farmers received the benefit. and we don't have pre-intervention data for comparison. can anybody suggest me robust methodology for impact evaluation without control group and baseline data.
thank you.
It is difficult to conduct impact study without baseline data.Some sorts of baseline data is required,otherwise before-after design is not possible.PRA tools may be utilized to get some baseline information quickly from the beneficiaries. However, following link may be useful:
Bamberger, M., (2010). Reconstructing Baseline Data for Impact Evaluation and Results Measurement, The World Bank. Retrieved from: http://siteresources.worldbank.org/INTPOVERTY/Resources/335642-1276521901256/premnoteME4.pdf
Contents lists available at SciVerse Science Direct Journal of Development Economics journal homepage: www.elsevier.com/locate/devec.
Beyond baseline and follow-up: The case for more T in experiments
☆ David McKenzie ⁎ World Bank, United States
Dear Kolla Sravanthi
In these circumstances you can base your evaluation on:
1. The theory of change - and test the links in the causal chain and assumptions
2. Secondary data, for instance compare data from the intervention area to the national data
3. Triangulation of data sources and methods - including mixed method interviews with stakeholders
This text might be useful: http://www.oecd.org/dac/evaluation/dcdndep/37671602.pdf
I would add one more option to Nyawira Mwangi. That is, use a pre-post design. While this does document change, it is subject to other explanations of causation such as influence by outside events and maturation of the subjects. However, it is relatively simple to do and may be useful with one of the three suggestions by Mwangi are not an option.
Without "baseline" data, I'm not sure that a pre-post design is possible.
Do you have independent variables that describe the participants? If so, then you could conduct a regression analysis where something like "level of adoption" would be the dependent variable.
It is difficult to conduct impact study without baseline data.Some sorts of baseline data is required,otherwise before-after design is not possible.PRA tools may be utilized to get some baseline information quickly from the beneficiaries. However, following link may be useful:
Bamberger, M., (2010). Reconstructing Baseline Data for Impact Evaluation and Results Measurement, The World Bank. Retrieved from: http://siteresources.worldbank.org/INTPOVERTY/Resources/335642-1276521901256/premnoteME4.pdf
Contents lists available at SciVerse Science Direct Journal of Development Economics journal homepage: www.elsevier.com/locate/devec.
Beyond baseline and follow-up: The case for more T in experiments
☆ David McKenzie ⁎ World Bank, United States
I agree with Nyawira Mwangi's response, and would emphasize the value of qualitative methods such as interviewing for documenting causal processes (as opposed to simply showing a pre-post difference). I've attached a paper that discusses this; interviews can substantially support the claim that cash transfers had an impact, although you need to get beyond simply asking whether they did so and seek concrete evidence for their effect.
- Badges
- Science method
- Similar topics
- Methodology