Difference between revisions of "Ieduplicates"

Jump to: navigation, search
(114 intermediate revisions by 6 users not shown)
Line 1: Line 1:
'''ieduplicates''' and the sister command '''iecompdup''' are used to identify and resolve duplicates in raw survey data.
<code>ieduplicates</code> is the second command in the Stata package created by [https://www.worldbank.org/en/research/dime/data-and-analytics DIME Analytics], <code>[[iefieldkit]]</code>. <code>ieduplicates</code> identifies [[Duplicates and Survey Logs | duplicate values]] in [[ID Variable Properties|ID variables]]. '''ID variables''' are variables that uniquely identify every [[Unit of Observation|observation]] in a dataset, for example, '''household_id'''. It then exports them to an Excel file that the [[Impact Evaluation Team|research team]] can use to resolve these duplicates. The '''research team''' should run <code>ieduplicates</code> with each new batch of incoming data to ensure [[Monitoring Data Quality|high quality data]] before [[Data Cleaning | cleaning]] and [[Data Analysis | analysis]].
==Read First==
* Please refer to [[Stata Coding Practices|Stata coding practices]] for coding best practices in Stata.
* <code>ieduplicates</code> is part of the package <code>[[iefieldkit]]</code>, which has been developed by [https://www.worldbank.org/en/research/dime/data-and-analytics DIME Analytics].
* <code>ieduplicates</code> identifies duplicates in [[ID Variable Properties|ID variables]], and then <code>[[iecompdup]]</code> resolves these issues.
* To install <code>ieduplicates</code>, as well as other commands in the <code>[[iefieldkit]]</code> package, type <syntaxhighlight lang="Stata" inline>ssc install iefieldkit</syntaxhighlight> in Stata, as this command is a part of the <code>[[iefieldkit]]</code> package.
* For instructions and available options, type <syntaxhighlight lang="Stata" inline>help ieduplicates</syntaxhighlight>.


This article is means to describe use cases, work flow and the reasoning used when developing the commands. For instructions on how to use the command specifically in Stata and for a complete list of the options available, see the help files by typing <code>help ieduplicates</code> or <code>help iecompdup</code> in Stata.
== Overview ==
The <code>ieduplicates</code> and <code>[[iecompdup]]</code> commands are meant to help [[Impact Evaluation Team|research teams]] deal with [[Duplicates and Survey Logs|duplicate observations]] in [[Primary Data Collection|primary data]]. These commands are designed to identify and resolve duplicate instances of an [[ID Variable Properties|ID variable]] in raw survey data, and ensure that each observation is [[ID Variable Properties#Property 1: Uniquely Identifying|uniquely]] and
[[ID Variable Properties#Property 2: Fully Identifying|fully identified]]. The commands combine four key tasks to resolve duplicate values:
# '''Identifying duplicate entries.'''
# '''Comparing observations with the same ID value.'''
# '''Tracking and documenting changes to the ID variable.'''
# '''Applying the necessary corrections to the data.'''
In any dataset, certain variables should be unique on purpose, to allow the '''research team''' to identify them during further [[Data Analysis|analysis]]. For example, suppose you select '''household_id''' as the unique ID variable. Now suppose you pick the observation with '''household_id'''= '''123456''' for [[Back Checks|back checks]], but the dataset has two observations with '''household_id''' = '''123456'''. In this case, it is important to resolve these duplicate observations before performing the '''back check'''.  


== Intended use cases ==
When you run <code>ieduplicates</code> for the first time, it will create a [[Ieduplicates#Duplicates Correction Template|duplicates correction template]]. This template will list all observations that contain duplicate values of an '''ID variable''' that should be unique. In the example above, after creating this template, <code>ieduplicates</code> will, by default, display a message pointing out that '''household_id''' does not '''uniquely''' and '''fully identify''' the data. It will also stop your code, and require you to fill the correction template before you can move on.
'''ieduplicates''' is meant to be used directly after importing raw data from for example a survey data collection. It does two things. It outputs a report of all the duplicated (the report can be used for correcting the duplicates) and it removes the duplicates from the data set until they are resolved. The reason the duplicates are resolved is to make sure that other quality checks that use the ID require unique IDs. For example, if household with ID 123456 was selected for back checks but you incorrectly have two observations that were given the ID 123456, then it is better to solve that duplicate first (you can use the report for this) before trying to run the back check test on either of the observations. Make sure not to overwrite the original raw data with the data set where ieduplicates has removed the duplicates.


'''iecompdup''' helps you to identify the reason for the duplicates. The decision on how to correct a duplicate is always a qualitative decision, but iecompdup compares the duplicated quantitatively and in almost all cases gives you the information that you need in order to make the qualitative decision.
== Syntax ==
The basic syntax for <code>ieduplicates</code> is as follows:
<syntaxhighlight lang="Stata">ieduplicates id_varname using "filename.xlsx"
    , uniquevars(varlist)</syntaxhighlight>
As inputs, <code>ieduplicates</code> requires the following :
* '''id_varname</code>''': This is the name of the single, unique [[ID Variable Properties|ID variable]]. This variable must be such that it would be an unacceptable duplicate in the dataset, and so must never be repeated. If there are two or more variables that identify the observation in the dataset, you should create a single '''ID variable''' that is unique for the dataset. This variable could be either a string or a number. For example, '''household_id'''.
* '''''"filename.xlsx"''''':  This provides the name of the Excel file in which <code>ieduplicates</code> will display the '''duplicates correction template'''. The file name in this case is specified with the help of <code>using</code>, and must include an [[Stata Coding Practices#File Paths|absolute file path]]. For example, '''"C:/myIE/Documentation/DupReport.xlsx"''' is the '''absolute file path''' for the file called '''"DupReport.xlsx"'''. Since the output is an Excel sheet, even those [[Impact Evaluation Team|members of the research team]] who do not know Stata can read the report, and make corrections.
* <code>uniquevars()</code>: Finally, <code>ieduplicates</code> uses one or multiple variables specified within '''<code>()</code>''' to uniquely identify each observation in the dataset. However, most [[Computer-Assisted_Personal_Interviews_(CAPI)#CAPI_Software|data collection tools]] only use one variable for this purpose. For example, [[SurveyCTO Coding Practices|SurveyCTO]] creates this variable automatically, and names it '''"KEY"'''.
For example, if there are no observations which have duplicate values of '''household_id''', for instance, <code>ieduplicates</code> will display a message saying the data set is '''uniquely''' and '''fully identified''' on the basis of '''household_id'''. In such a case, there will be no output, and this command will leave the data unchanged.  


=== Intended Work Flow ===
However, if there are observations which have duplicate values of '''household_id''', <code>ieduplicates</code> will save the output to an Excel sheet called '''"DupReport.xlsx"'''. This file will contain information on these observations in the form of the '''duplicates correction template''', and <code>ieduplicates</code> will also stop your code with a message listing the repeated values under '''household_id'''. See '''Figure 1''' below for a sample correction template.


# Run '''ieduplicates''' on the raw data
== Implementation ==
#* If there are no duplicates, then you are done and can skip the rest of this list.  
The following is the process for using <code>ieduplicates</code>:
# If there are duplicates, use '''iecompdup''' on any duplicates identified.  
# '''Run''' <code>ieduplicates</code> ''' on the raw data.''' If there are no duplicate observations, then you are done. Skip the rest of the steps. If there are duplicates, the command will output an Excel file which contains the '''duplicates correction template'''. It will display a message with a link to this file, and stop the code from moving forward. It will also show a message listing the duplicate ID values. <br>
# Enter the corrections identifies with '''iecompdup''' to the duplicates in the report outputted by '''ieduplicates'''
# '''Open the duplicates correction template.''' This template will list each duplicate entry of the ID variable, and information about each observation. It also contains 5 blank columns - '''correct''', '''drop''', '''newid''', '''initials''', and '''notes'''. Use these columns to make corrections, and include comments to [[Data Documentation|document]] the corrections. <br>
# After entering the corrections, save the report in the same location with the same name,
# '''Use''' <code>[[iecompdup]]</code> '''for more information.''' Sometimes the template is not enough to solve a particular issue. In such cases, run the <code>[[iecompdup]]</code> command on the same dataset. <br>
# Run '''ieduplicates''' again. The corrections you have entered is now applied and only duplicates that are still not resolved are removed this times.
# '''Overwrite the previous file.''' After entering all the corrections to the template, save the Excel file in the same location with the same name. <br>
# '''Run''' <code>ieduplicates</code> '''again.''' This will apply the corrections you made in the previous steps. Now if you use the '''force''' option, it will only remove those duplicates that you did not resolve. <br>
# '''Do not overwrite the orginal raw data.''' Save the resulting dataset under a different [[Naming Conventions|name]].<br>
# '''Repeat these steps with each new round of data.'''
==== Sample template ====
A sample '''duplicates correction template''' created by <code>ieduplicates</code> is displayed below.
[[File:Ieduplicates.png|900px|thumb|center|'''Figure 1: Duplicates Correction Template''']]


Repeat these steps every time you get new data. Our recommendation is that this is done every day that you have new data.
== Duplicates Correction Template ==
The <code>ieduplicates</code> exports the '''duplicates correction template''' to the Excel file, based on the following syntax:
<syntaxhighlight lang="Stata">ieduplicates id_varname using "filename.xlsx"
    , uniquevars(varlist)
    [force keepvars(varlist) tostringok droprest nodaily
    duplistid(string) datelisted(string) datefixed(string) correct(string)
    drop(string) newid(string) initials(''string) notes(string) listofdiffs(string)]</syntaxhighlight>


== Directions ==
* <code>force</code>: Removes all observations that contain duplicate values of '''id_varname''' from the data. As a result, it keeps only '''uniquely''' and '''fully identified''' observations. The <code>force</code> option is required so that you know that <code>ieduplicates</code> is making changes to your dataset. Do not overwrite the original raw data with the one that this command generates, otherwise you will lose the original data. Save this new dataset with a different name.


== Reasoning used during development ==
* <code>duplistid</code>: Uniquely identifies each of the duplicate observations in the '''duplicates correction template'''. It does this by assigning '''1''' to the first instance of a duplicate value, '''2''' to the second instance, and so on.
 
* <code>datelisted</code>: Indicates the date on which the observation was first included in the template.
 
* <code>listofdiffs</code>: Lists the variables in the data set that are different for the observations that have duplicate values of the ID variable. You can also rename these columns by specifying the new column name under their respective options.
 
* <code>correct</code>, <code>drop</code>, <code>newID</code>, <code>initials</code>, and <code>notes</code>: Creates 5 blank columns which you can fill to make corrections where needed, and complete the template.
 
This completed template then acts like a permanent [[Data Documentation|documentation]] of how the [[Impact Evaluation Team|research team]] resolved duplicate ID variables in the raw data. There are three options for
resolving duplicate observations. They appear in the form of '''''correct''''', '''''drop''''', and '''''newID''''' columns in the template. Consider the following examples to understand how to fill these columns:
* '''''drop''''': If you want to keep one of the duplicate observations and drop another, then write '''“correct”''' in the '''''correct''''' column for the observation you want to keep, and '''“drop”''' in the '''''drop''''' column for the observation you want to drop. Make sure you mention the correct value under the  '''''key''''' column for the observations you want to keep, and the ones that you want to drop from the dataset. This is important because '''SurveyCTO''' creates a unique '''key''' for every observation, even if two or more observations have duplicate variable IDs by mistake.
 
* '''''newID''''': If you want to keep one of the duplicates and assign a new unique ID to another one, write '''“correct”''' in the '''''correct''''' column for the observation you want to keep, and the new corrected ID value in the '''''newID''''' column for the observation to which you want to assign the new unique ID.
 
* '''Combine both methods''': You can also combine these two methods if you have more than 2 duplicate observations. Note that you must always indicate which observation you want to keep for each group of duplicate observations.
 
After entering your corrections, save the file and run <code>ieduplicates</code> again to apply the corrections to the dataset.
Since <code>ieduplicates</code> should be used frequently as new data comes in from the field, the command also manages a subfolder called '''/Daily/''' in the same folder which contains the main Excel file. <code>ieduplicates</code> uses this subfolder to save a backup version (along with the date) for every time the template is updated. If you do not wish to use this feature, use the '''nodaily''' option which prevents the creation of backups.
 
'''Note:''' If two different templates are generated on the same day, it saves the second with an additional time stamp on the name. This is especially useful in case the main corrections template, or any of its contents get deleted. You can restore a backup version by simply copying it out of the '''/Daily/''' folder and remove the date from the name.
 
== Related Pages ==
[[Special:WhatLinksHere/Ieduplicates|Click here for pages that link to this topic.]]<br>
This page is part of the topic <code>[[iefieldkit]]</code>. Also see <code>[[iecompdup]]</code>.
 
== Additional Resources ==
* DIME Analytics (World Bank), [https://osf.io/uc2en/ Real Time Data Quality Checks]
* DIME Analytics (World Bank), [https://github.com/worldbank/iefieldkit The <code>iefieldkit</code> GitHub page]
[[Category: Stata ]]

Revision as of 22:36, 18 November 2020

ieduplicates is the second command in the Stata package created by DIME Analytics, iefieldkit. ieduplicates identifies duplicate values in ID variables. ID variables are variables that uniquely identify every observation in a dataset, for example, household_id. It then exports them to an Excel file that the research team can use to resolve these duplicates. The research team should run ieduplicates with each new batch of incoming data to ensure high quality data before cleaning and analysis.

Read First

  • Please refer to Stata coding practices for coding best practices in Stata.
  • ieduplicates is part of the package iefieldkit, which has been developed by DIME Analytics.
  • ieduplicates identifies duplicates in ID variables, and then iecompdup resolves these issues.
  • To install ieduplicates, as well as other commands in the iefieldkit package, type ssc install iefieldkit in Stata, as this command is a part of the iefieldkit package.
  • For instructions and available options, type help ieduplicates.

Overview

The ieduplicates and iecompdup commands are meant to help research teams deal with duplicate observations in primary data. These commands are designed to identify and resolve duplicate instances of an ID variable in raw survey data, and ensure that each observation is uniquely and fully identified. The commands combine four key tasks to resolve duplicate values:

  1. Identifying duplicate entries.
  2. Comparing observations with the same ID value.
  3. Tracking and documenting changes to the ID variable.
  4. Applying the necessary corrections to the data.

In any dataset, certain variables should be unique on purpose, to allow the research team to identify them during further analysis. For example, suppose you select household_id as the unique ID variable. Now suppose you pick the observation with household_id= 123456 for back checks, but the dataset has two observations with household_id = 123456. In this case, it is important to resolve these duplicate observations before performing the back check.

When you run ieduplicates for the first time, it will create a duplicates correction template. This template will list all observations that contain duplicate values of an ID variable that should be unique. In the example above, after creating this template, ieduplicates will, by default, display a message pointing out that household_id does not uniquely and fully identify the data. It will also stop your code, and require you to fill the correction template before you can move on.

Syntax

The basic syntax for ieduplicates is as follows:

ieduplicates id_varname using "filename.xlsx"
     , uniquevars(varlist)

As inputs, ieduplicates requires the following :

  • id_varname: This is the name of the single, unique ID variable. This variable must be such that it would be an unacceptable duplicate in the dataset, and so must never be repeated. If there are two or more variables that identify the observation in the dataset, you should create a single ID variable that is unique for the dataset. This variable could be either a string or a number. For example, household_id.
  • "filename.xlsx": This provides the name of the Excel file in which ieduplicates will display the duplicates correction template. The file name in this case is specified with the help of using, and must include an absolute file path. For example, "C:/myIE/Documentation/DupReport.xlsx" is the absolute file path for the file called "DupReport.xlsx". Since the output is an Excel sheet, even those members of the research team who do not know Stata can read the report, and make corrections.
  • uniquevars(): Finally, ieduplicates uses one or multiple variables specified within () to uniquely identify each observation in the dataset. However, most data collection tools only use one variable for this purpose. For example, SurveyCTO creates this variable automatically, and names it "KEY".

For example, if there are no observations which have duplicate values of household_id, for instance, ieduplicates will display a message saying the data set is uniquely and fully identified on the basis of household_id. In such a case, there will be no output, and this command will leave the data unchanged.

However, if there are observations which have duplicate values of household_id, ieduplicates will save the output to an Excel sheet called "DupReport.xlsx". This file will contain information on these observations in the form of the duplicates correction template, and ieduplicates will also stop your code with a message listing the repeated values under household_id. See Figure 1 below for a sample correction template.

Implementation

The following is the process for using ieduplicates:

  1. Run ieduplicates on the raw data. If there are no duplicate observations, then you are done. Skip the rest of the steps. If there are duplicates, the command will output an Excel file which contains the duplicates correction template. It will display a message with a link to this file, and stop the code from moving forward. It will also show a message listing the duplicate ID values.
  2. Open the duplicates correction template. This template will list each duplicate entry of the ID variable, and information about each observation. It also contains 5 blank columns - correct, drop, newid, initials, and notes. Use these columns to make corrections, and include comments to document the corrections.
  3. Use iecompdup for more information. Sometimes the template is not enough to solve a particular issue. In such cases, run the iecompdup command on the same dataset.
  4. Overwrite the previous file. After entering all the corrections to the template, save the Excel file in the same location with the same name.
  5. Run ieduplicates again. This will apply the corrections you made in the previous steps. Now if you use the force option, it will only remove those duplicates that you did not resolve.
  6. Do not overwrite the orginal raw data. Save the resulting dataset under a different name.
  7. Repeat these steps with each new round of data.

Sample template

A sample duplicates correction template created by ieduplicates is displayed below.

Figure 1: Duplicates Correction Template

Duplicates Correction Template

The ieduplicates exports the duplicates correction template to the Excel file, based on the following syntax:

ieduplicates id_varname using "filename.xlsx" 
     , uniquevars(varlist)
     [force keepvars(varlist) tostringok droprest nodaily 
     duplistid(string) datelisted(string) datefixed(string) correct(string)
     drop(string) newid(string) initials(''string) notes(string) listofdiffs(string)]
  • force: Removes all observations that contain duplicate values of id_varname from the data. As a result, it keeps only uniquely and fully identified observations. The force option is required so that you know that ieduplicates is making changes to your dataset. Do not overwrite the original raw data with the one that this command generates, otherwise you will lose the original data. Save this new dataset with a different name.
  • duplistid: Uniquely identifies each of the duplicate observations in the duplicates correction template. It does this by assigning 1 to the first instance of a duplicate value, 2 to the second instance, and so on.
  • datelisted: Indicates the date on which the observation was first included in the template.
  • listofdiffs: Lists the variables in the data set that are different for the observations that have duplicate values of the ID variable. You can also rename these columns by specifying the new column name under their respective options.
  • correct, drop, newID, initials, and notes: Creates 5 blank columns which you can fill to make corrections where needed, and complete the template.

This completed template then acts like a permanent documentation of how the research team resolved duplicate ID variables in the raw data. There are three options for resolving duplicate observations. They appear in the form of correct, drop, and newID columns in the template. Consider the following examples to understand how to fill these columns:

  • drop: If you want to keep one of the duplicate observations and drop another, then write “correct” in the correct column for the observation you want to keep, and “drop” in the drop column for the observation you want to drop. Make sure you mention the correct value under the key column for the observations you want to keep, and the ones that you want to drop from the dataset. This is important because SurveyCTO creates a unique key for every observation, even if two or more observations have duplicate variable IDs by mistake.
  • newID: If you want to keep one of the duplicates and assign a new unique ID to another one, write “correct” in the correct column for the observation you want to keep, and the new corrected ID value in the newID column for the observation to which you want to assign the new unique ID.
  • Combine both methods: You can also combine these two methods if you have more than 2 duplicate observations. Note that you must always indicate which observation you want to keep for each group of duplicate observations.

After entering your corrections, save the file and run ieduplicates again to apply the corrections to the dataset. Since ieduplicates should be used frequently as new data comes in from the field, the command also manages a subfolder called /Daily/ in the same folder which contains the main Excel file. ieduplicates uses this subfolder to save a backup version (along with the date) for every time the template is updated. If you do not wish to use this feature, use the nodaily option which prevents the creation of backups.

Note: If two different templates are generated on the same day, it saves the second with an additional time stamp on the name. This is especially useful in case the main corrections template, or any of its contents get deleted. You can restore a backup version by simply copying it out of the /Daily/ folder and remove the date from the name.

Related Pages

Click here for pages that link to this topic.
This page is part of the topic iefieldkit. Also see iecompdup.

Additional Resources