Cash Transfer Post Distribution Monitoring (PDM) Guide 1. Methodology 1.1 Focus Group Discussion How to select community According to the discussion with cash transfer M&E technical working group, X {at least 2 clusters per district} clusters/community per district will be selected in every month to conduct focus group discussion immediately after cash transfer. The number of clusters/community is decided by considering the time and cost to be spent. Two/Three focus group discussions (male and female; children) will be conducted in each cluster/community. Each cluster comprises at least 8-12 members randomly selected from beneficiary list. How to select the group member: Systematic sampling which is a modification of simple random sampling that picks every nth household from list of households will be used for group members for FGD. First take distribution list of selected community sort by alphabetical order. Then calculate cumulative population. Use the following steps: List down all beneficiaries of community; Calculate the cumulative population by adding each beneficiary; Identify sample interval as follows n = N/S Where, n= Sample interval N= total population S= Sample size Take a random number between one and the sample interval The beneficiary with cumulative population number that fall in the selected random number is the first selected sample; By adding the sample interval on the cumulative population size of each selected sample all the rest samples are selected. Example: Beneficary Cum. list Population Pop. Sample Remark Fatu 1 1 Hawa 1 2 x Mary 1 3 George W. 1 4 1 Amadu 1 . 1 Leolee 5 10 X . . Total 1 100 100 n= 100/12= 8 Let us take sample interval 2, (which between 1 and 8). Therefore, our firs sample is Hawa. Then 2+8 =10, the second sample, third sample will be 10+8=18 so on. Based on the above sampling method we can get 12 members of group participants. Using FGD checklist we can interview each group and summarize the responses for further compilation and write the report based on the reporting outline. Who will participate in FGD A HH member who is eligible for FGD is the first recipient of cash transfer; a mother or any care giver in the HH (even the father) who have direct involvement of the cash transfer. 1.2 Quantitative Survey Survey Methodology Two stages Cluster sampling method is going to be employed in the survey. Stage one is community and (make first cluster equally representative) and stage two is list of beneficiary randomly selected from beneficiary list. A total of X households from 30 clusters (community) are going to be interviewed. Sample Size Indicators may be expressed either as a proportion, a mean, or a total. The following formula (Basic Equation 1) may be used to calculate the required sample size for indicators expressed as a percentage or proportion for comparative studies. 2 n = D * (Zα + Zβ)2 * [ p1 (1-p1) + p2 (1-p2)] / [(p2 – p1)2 * (1-r)] n = minimum sample size required D = design effect; D = 2 if you use two-stage cluster sampling design p1 = estimated value of the key indicator at baseline (or control group) p2 = expected value of the key indicator at the end-line (or intervention) Zα = Z-score for the desired degree of confidence (α = .95, Zα =1.645) Zβ = Z-score for the desired level of power (β = 0.80. thus Zβ =0.84) r = non-response error (normally = 0.1, 10%) Formulae for Sample Size If target population is small, say N < 15,000 HHs, use the correction factor to calculate an adjusted sample size, say n. n= n0 _____ = 1 + (n0 – 1) / N n0 = n0______ 1 + n0 / N t2 p q_____ d2 n= Sample size d= level of error ready to accept on the estimate (5%= 0.05 or 10% = 0.10) t = z-value for confidence levels (95%= 1.96 or 90%=1.64) N = Population size q= 1-p Sample size is maximum if p=0.5, normally use this assumption if you are not sure about the guessed value of the estimate of p. 3 Sampling procedure and sample size for interview -Selection of clusters For this survey, WHO 30 X 30 cluster method will be adopted. Based on sample size determined, X number of HHs per cluster will be interviewed. Cluster here is refer to community. Community by county will be determined using population proportion sampling method. And then from list of all community in each county Y number of community is selected randomly (lottery method). Selection of Households After selecting the cluster/community/ based on lottery method, we have to list down all beneficiary households of each cluster/community/ in alphabetical order. Then systematic sampling which is a modification of simple random sampling that picks every nth household from list of households is going to be applied for selecting the sample households. First take beneficiary list of each cluster/community/ and then calculate cumulative population based on the following formula. Use the following steps: List down all beneficiaries of the cluster/community/ in alphabetic order; Calculate the cumulative population by adding up each beneficiary; Identify sample interval as follows n = N/S Where, n= Sample interval N= total population S= Sample size Take a random number between one and the sample interval The beneficiary with cumulative population number that fall in the selected random number is the first selected sample; By adding the sample interval on the cumulative population size of each selected sample all the rest samples are selected. Example: Beneficary Populatio list n Cum. Pop. Sampled HH Remark Annie 1 1 4 Arthur Augustine Fedrick W. 1 1 1 James John Johnson Joseph Joseph 1 1 1 1 1 Kou Marie Martha Martha Martha Marvin Marvin Mary Mary Oldma 1 1 1 1 1 1 1 1 1 1 Total 2 3 4 . . 11 12 13 14 15 . . 21 22 23 24 25 26 27 28 29 30 200 X X 200 n= 200/10= 20 Let us take first random sample 3, (which between 1 and 20). Therefore, our first sample is Augustine. Then 3+20 =23, the second sample will be Martha, and third sample will be 23+20=43 so on. Continue the process of selection until the required sample household is chosen. Who will be interviewed? A HH member who is eligible for interview is the first recipient of cash transfer; a mother or any care giver in the HH (even the father) who have direct involvement of the cash transfer. 5 6