Adeline Delavande, Xavier Giné, and David McKenzie, "Eliciting Probabilistic Expectations with Visual Aids in Developing Countries: How Sensitive Are Answers to Variations in Elicitation Design?", Journal of Applied Econometrics, Vol. 26, No. 3, 2011, pp. 479-497. The file questionnaires.zip contains 8 versions of the questionnaire in PDF format. The file dgm-data.zip contains three data files and a Stata do file. These files are ASCII files in DOS format, so Unix/Linux users should use "unzip -a". 1. data file: fish_dat: 270 observations and 60 variables This file contains the answers from the questionnaire. id Boat Owner ID version version of questionnaire used (randomization and ordering of question) C denotes Common support I denotes Individual support max: when individual support is elicited, the maximum is elicited first min: when individual support is elicited, the minimum is elicited first 10: 10 beans are used 20: 20 beans are used. For example, "IminC20" means that first the distribution with individual support is elicited, and when doing so, the min is first asked and then the max. Finally, the respondent will be asked to use 20 stones. bo_name Boat Owner Name village_name Village Name day Day of interview month Month of interview q1 question 1 q2 question 2 q3 question 3 q4 question 4 q5 question 5 aug_catchdist1 to aug_catchdist20 : subjective distribution of catches one day in August elicited with common support. The answer in aug_catchdistX indicates the number of beans that respondents allocated in the Xth intervals. min_earn the MINIMUM you expect to make that day with the BOAT you own p25 2nd threshold in the individual support midpoint 3rd threshold in the individual support p75 4th threshold in the individual support max_earn the MAXIMUM you expect to make that day with the BOAT you own aug_catchrange1 to aug_catchrange4: subjective distribution of catches one day in August elicited with individual support. The answer in aug_catchrangeX indicates the number of beans that respondents allocated in the Xth intervals. q10_01 to q10_20: subjective distribution of catches one day in August for another Boat-owner. The answer in q10_X indicates the number of beans that respondents allocated in the Xth intervals. 2. data file: catches_all, 8,339 observations, 4 variables. This file provides the realized catches data for the month of August. Each row is the realization of the sales for one boat owner on the day specified by date. VID Village id id Boat Owner ID date Date of the sales sales Sales amount Note: sales equal to zero indicates that they did not fish. To get individual means of sales for the month of August, one should follow the following steps: * Drop all zero catches since this means they didn't fish; drop if sales==0 * Generate mean by Boatowner; egen meanactual=mean(sales), by(id) 3. data file: bo_char, 272 observations, 8 variables This file provides the demographic characteristics of the boat owners. id Boat Owner ID age age edu years of education lit literate nchild number of children years_BO years as boat owner num_fb number of fiber boat ncrew number of crew 4. Boatexpv4.do: Stata do-file to organize the data for analysis. Adeline Delavande, Universidade Nova de Lisboa and RAND Xavier Giné, World Bank and BREAD David McKenzie, World Bank, BREAD and IZA