MSD: introductory videos
What is a market system?
Why a systems approach?
Features of a market systems approach
M4P Operational Guide
Evidence Review 2019
BEAM Evidence Map
About the Evidence Map
DCED Evidence Framework
Monitoring tips and techniques
How to! Notes
What is the MSD Competency Framework?
Guidance for mentors and managers
Guidance for hiring and recruitment
Personal development for practitioners
Events & training
Statistical tool-market analysis and programmatic decision
Reply on DGroups
| 2 comments
Applying statistical tools in market analyses and CVA programmatic decisions
Dec. 23, 2020, 4:33 p.m.
In my experience, the type of market analysis typically used for CVA programmatic decisions (often using tools like RAM, EMMA, 48 Hour Tool, MSMA Toolkit) is poorly suited for statistical analysis of any kind, for a couple main reasons:
* Sample sizes tend to be very small (and in some cases they are by definition small - e.g. there is often only a handful of wholesalers in any given market)
* The data gathered during these assessments is approximate, not precise. Many vendors in the types of markets we work in don't keep perfect record books and/or aren't willing to share them with unknown enumerators, even if they do. For this reason we often encourage people to ask for ranges of figures from vendors rather than exact numbers (e.g. I sell anywhere from 50-80 kg of maize flour per week instead of I sell an average of 68 kg of maize flour per week). Estimates and ranges don't lend themselves to statistical analysis.
The good news is that we seem to be able to get away with this sort of "good enough", unscientific analysis, in that it tends to be sufficient to inform the kinds of decisions we need to make for humanitarian programming. The EMMA Toolkit <https: 447="" download?token="QFRt_6Ir" file="" www.emma-toolkit.org="">, especially the section on Market Analysis <https: bundle="" default="" files="" sites="" step8.pdf="" www.emma-toolkit.org="">, has a pretty good explanation of how this can work.
I know that WFP often works with larger data sets than other NGOs do, so someone from WFP might have a different take on this.
Dec. 23, 2020, 1:55 p.m.
I am exploring a possibility of applying statistical tools in market analyses and CVA programmatic decisions. My question is:
* Is there any experience of using bivariate (e.g. T-test/Chi-Square test) or multivariate tools (F-test, Z test etc) to support market analyses and programmatic decision? If yes, which one, and what significance level (or confidence interval)? Any further advice on analysis of variance or regression analysis would also be very useful.
[https://plan-international.org/sign-the-letter?utm_source=external_banner&utm_medium=email&utm_campaign=ggefo] <https: plan-international.org="" sign-the-letter?utm_source="external_banner&utm_medium=email&utm_campaign=ggefo">
English Campaign Tag
Any opinions expressed in this message and any attachments are those of the sender only and do not necessarily represent the views of Plan. Internet communications are not secure and Plan accepts no responsibility for the content of this e-mail.
The information contained in this message and any attachments is intended solely for the use of the person(s) to whom the message is addressed. The information may be confidential and, if you are not the intended recipient, you must not copy, distribute or take any action in relation to it.
For the content of this e-mail to be contractually binding, it must be signed by an authorised representative of Plan.
A Limited Company Registered in England No. 03001663.
Registered Office: Dukes Court, Duke Street, Woking, GU21 5BH, United Kingdom
Plan Limited is a wholly-owned subsidiary of Plan International, Inc. (a not-for-profit corporation registered in New York State, USA) DF
Plan Disclaimer Tag</https:>
on 06/20/2021 at 14:57