Difference between revisions of "DataWork Folder"

Jump to: navigation, search
Line 1: Line 1:
A well organized data folder reduces the risk for many types of errors. At DIME, we have a standardized folder structure. Some projects have special folder requirements and only use the folder set up as a starting point, but many resources created by DIME are easier to take advantage of if this template is followed. It takes a lot of work to reorganize a project folder, so we strongly recommend that projects follow our standard from the beginning. A poorly set up folder will have inefficiency consequences and increases the risk of errors over several years.
A well organized data folder reduces the risk for many types of errors. At DIME, we have a standardized folder structure. Some projects have special folder requirements and only use the folder set up as a starting point, but many resources created by DIME are easier to take advantage of if this template is followed. It takes a lot of work to reorganize a project folder, so we strongly recommend that projects follow our standard from the beginning. A poorly set up folder will have inefficiency consequences and increases the risk of errors over several years.


We have published a command called [[iefolder]] in our package [[Stata_Coding_Practices#ietoolkit|ietoolkit]] that we have published on SSC. <code>iefolder</code> sets up the recommended folder structure described in this article for you.
We have published a command called <code>[[iefolder]]</code> in our package [[Stata_Coding_Practices#ietoolkit|<code>ietoolkit</code>]] that we have published on SSC. <code>iefolder</code> sets up the recommended folder structure described in this article for you.


== Read First ==
== Read First ==

Revision as of 20:04, 27 March 2017

A well organized data folder reduces the risk for many types of errors. At DIME, we have a standardized folder structure. Some projects have special folder requirements and only use the folder set up as a starting point, but many resources created by DIME are easier to take advantage of if this template is followed. It takes a lot of work to reorganize a project folder, so we strongly recommend that projects follow our standard from the beginning. A poorly set up folder will have inefficiency consequences and increases the risk of errors over several years.

We have published a command called iefolder in our package ietoolkit that we have published on SSC. iefolder sets up the recommended folder structure described in this article for you.

Read First

  • Do not set up these folders manually. iefolder is a Stata command that easily sets up and updates this folder structure for you

Where should the DataWork folder be created?

Image 1. Example of where the DataWork folder is location in relation to Box/DropBox folders. (Click to enlarge.)

Most folders are shared across the project teams using a DropBox, Box or similar. In this folder there are usually a lot of folders for project budget, government communications etc. The DataWork folder is assumed to be one of them.

See the Image 1 to the right with one example of a Box/DropBox folder with three project folders. All three projects has a similar sub-folder structure, but in the image only one of the projects sub-folder structure is show. The DataWork folder is highlighted with a red circle.

Anything related to the data of a project has a designated location inside this folder. This includes data-files, sampling and treatment assignment code, questionnaires, data collection documentation, analysis code, analysis output etc. This includes data collected by hour selves, both regular survey rounds and monitor data, but it should also include other sources of data such as admin data or secondary data.

Inside the DataWork folder

File:FolderDataWork.png
Image 2. Example of a DataWork folder. (Click to enlarge.)

Inside the DataWork folder there should only be folders and files that help navigating those folders. In our standardized folder there are only three types of folders; Survey Rounds (Baseline, Endline etc.), Monitor Data and Master Data Sets. What's described here is only a template structure and additional folders are often needed. But we recommend that you try to fit even additional folders into one of these folder types and create them using iefolder.

In addition to the folders, in our standardized folder structure, there is also a Project_MasterDofile inside the DataWork folder. This file have three purposes. The first two are described in detail in the article for Master Dofiles, but in short they are that it makes it possible to run all code related to one project at the same time, and it also sets up all the folder paths required to run any dofile for this project. The third purpose is that this file is the main map to the DataWork folder. Since all code can be run from this file, and since all outputs are (indeirectly) created by this file, this file is the starting point to find where any do-file, data set or output is located in the DataWork folder. Another examples of files that helps with the navigation of the folder could be a Word document or a PDF describing how to navigate the sub-folders. Such files are not included in our folder template, but may sometimes be a good addition. Although, keep the number of files in this folder to an absolute minimum.

Survey Round

Image 3. Example of a Survey Round folder. (Click to enlarge.)

Baselines, Follow Up Surveys, Midlines, Endlines are examples of a Survey Round. This is the data that we in Impact Evaluations will test if it changes over time and if that change is significantly different between treatment and control. In contrast, the information in the master data sets, like the ID assigned by us, weather you were sampled for baseline, weather you are selected for treatment or control are all examples of information that is time invariant and will remain the same over the course of the project. Monitor data might change over time, for example in a impact evaluation running over many years one observation might not take up the treatment the first year but might do so the next year.

Each survey round should have it's own sub-folder inside the DataWork folder. For example - Inside the main data folder, you can have sub-folders like baseline, follow up 1, follow up 2, midline, endline, etc. See image 3 for an example. When you create a survey round folder using the command iefolder all sub-folders and sub-sub-folders described below will be created for you and all your master dofiles will be updated or created accordingly.

Inside each Survey Round folder you will find a master do file for that survey round as well as the following folders DataSets, Dofiles, Outputs, Documentation, and Questionnaire.

Multiple Units of Observation

If you have multiple units of observation in a survey round for example farmers and villages, or students, teachers and schools, then you should create a survey round folder for each unit of observations. For example, you would end up with one survey round folder called baseline_students, one called baseline_teachers and one called baseline_schools.

Sampling and Treatment Assignment

Sampling and Treatment Assignment folders have intentionally been left out from the survey round folders. Separate folders for those tasks has been created in the master data set folder as we strongly recommend that sampling and treatment assignment is done directly from the master data sets.

MonitorData

Monitor data is data collected to understand the implementation of the assigned treatment in the field. Survey round data helps us understand any changes in outcome variables that the treatment and other factors have caused during the duration of our project, and monitor data helps us the treatment its self. For example, who actually received the treatment and was the treatment carried out according to the research design. Monitor data helps us understand what is usually referred to as internal validity.

Since the purpose of survey round data and monitor data is slightly different, we recommend to keep the different types of data in different folders. There are sometimes overlap in survey round data and monitor data, and there it is not always feasible to keep them separated. The method used data collected as monitor data varies with each project. It can be both survey data, observation data and admin data provided by our partners. Monitoring data and survey round data should be kept separate unless the monitoring data was a part of the survey round interviews. For example, if an enumerator visited and training at the time of a midline survey and recorded attendance then that data should be kept separate from the survey round data. But if the respondents were asked in the a survey round survey if the respondent attended the training, then it not always worth the effort to go through the work of separating the monitor data from the survey data.

Monitor data has the same sub-folder structure as a survey round folder. The only difference in a folder structure perspective is that they are all organized as sub-folder to a folder in DataWork called MonitorData.

Admin Data

Admin data is data that have been collected for other purposes but has been shared with the research team. It can be used both in the way survey round data is used in the analysis or as the way monitoring data can be used. For example, if the outcome variable the research team is trying to evaluate is measured in some other way then that is admin data we can use to measure change in outcome data. One example would be standardized test scores. We can often use the internal data our implementation partner uses as monitor data. That is one example when admin data is monitoring data. There are also many cases when the same admin data set is used for both purposes.

We recommend that the project team classify admin data sets as either a survey round data or monitoring data and organize them in the DataWork folder accordingly.

MasterData

Image 6. Example of a Master Data Set folder. (Click to enlarge.)

Master Data Sets is the best tool to avoid errors related to using multiple sources of data for one project. All impact evaluations use different sources. Multiple survey rounds are in this sense different sources of data. A census made before a data collection is a different source of data. Monitoring data and admin data combined with any other data are other examples. Master data sets should therefore be used in all projects and iefolder therefore creates that folder by default when the DataWork folder is created.

The Master Data Sets is a listing of all observations we have ever encountered in relation to this project (not just the observations we sampled). In this listing we keep identifying information and the unique ID that we have assigned. We also keep time invariant information required at multiple stages of the research work. Examples are dummy for being sampled in baseline, categorical variable for treatment arm, dummy for correctly receiving the assign treatment arm etc. See the Master Data Sets article for more details.

Master Data Sets is more than just a folder. It is a methodology for internal research validity in a multi data set environment. See the Master Data Sets article for information both on how the master data set folder should be organized but also how it should be used for research quality purposes. One important motto is that if you do any merge (string or numeric) where the variable you merge on is not a proper ID variable, then one of the data set you are merging should always be a master data set.

Sampling and Treatment Assignment

The master data sets folder should also include all activities activities that is best practice to do directly on the main listing of all observations. The two best examples of such tasks in the context of Impact Evaluation is sampling and treatment assignment. This should never be done directly on census data or baseline data etc. When we have census data we use that data for sampling, but the point here is that we always want to match the census data to the master data and check that it makes sense in relation to whatever data we have there already. After that data quality step we can sample directly from the master data set knowing that the sample we randomize will make sense in relation to other data sources.

Back to Parent

This article is part of the topic Data Management