In validmeasures/myanmarMCCTdata: Myanmar Mother and Child Cash Transfer (MCCT) Programme Evaluation Data Handler

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)

myanmarMCCTdata: Myanmar Mother and Child Cash Transfer (MCCT) Programme Evaluation Data Handler

In support of the Myanmar Mother and Child Cash Transfer programme evaluation, this package has been developed to support the handling, processing, analysis of data collected for the study. The package includes functions to retrieve data from the online server database, to appropriately structure datasets, to check and clean data, to recode data to respective indicator sets, to estimate indicators and to perform appropriate comparative analysis specified by the analysis.

This package has been developed for use with R language for statistical computing. R can be downloaded freely from here and installed on Windows, MacOs and various Linux distributions. Guidance and support in installing R is available from the Comprehensive R Archive Network (CRAN).

Installation

Once you have installed R on your system, you can install myanmarMCCTdata from GitHub with the following commands in R:

if(!require(remotes) install.packages("remotes")
remotes::install_github("validmeasures/myanmarMCCTdata")

Usage

The myanmarMCCTdata package includes the following families of functions: 1) data retrieval and handling functions; 2) data cleaning functions; 3) data manipulation and structuring functions; 4) data recoding functions; and, 5) data analysis functions.

Data retrieval and handling functions

Data collected for the Myanmar MCCT Programme Evaluation is stored in an ODK Aggregate server hosted via ONA. Access to this server and to the data stored is restricted to those with explicit permissions and require a username and password.

Those with access can retrieve the Myanmar MCCT Programme Evaluation data from within R without having to manually download the data from the ONA server. This is facilitated by the get_mcct_data() function included in this package. To use this function, one requires the following information:

1. form ID - form/s are given unique identifiers. For the Myanmar MCCT Programme Evaluation, each survey form developed in ODK was given either of these unique identifiers:

   • baseline_mcct_final - main survey form
   • baseline_mcct_anthro_final - child and mother anthropometry form
   • baseline_townshipprofile_final - township profile form
   • baseline_villprofile_final - village profile form

2. username - this is the ODK Aggregate username for the person/organisation with permission to access and retrieve data from the server.

3. password - this is the corresponding ODK password for the person/organisation with permission to access and retrieve data from the server.

4. starting date and end date - Depending on whether you want to retrieve all the data or just part of the data, you will have to specify the starting date of the survey and the end date of the survey. If the data collection is still on-going, you may opt not to specify an end date and the current date will be used.

Given these, data for the main form of the Myanmar MCCT Programme Evaluation can be retrieved via the following commands in R:

surveyData <- get_mcct_data(id = "baseline_mcct_final",
                            start = "2019-08-06",
                            username = "ENTER_USERNAME_HERE",
                            password = "ENTER_PASSWORD_HERE")
````

A similar command is used to retrieve the data from the other forms used in the Myanmar MCCT Programme Evaluation:

```r
surveyData <- get_mcct_data(id = "baseline_mcct_anthro_final",
                            start = "2019-08-06",
                            username = "ENTER_USERNAME_HERE",
                            password = "ENTER_PASSWORD_HERE")
````

It should be noted that the commands above only retrieves the non-repeat components of the specified forms. If these forms have repeat components, these repeat components can be retrieved as well by specifying the argument `rep` as TRUE and then indicating the name/s of the repeating component/s in the argument `rep.names` as follows:

```r
surveyData <- get_mcct_data(id = "baseline_mcct_final",
                            start = "2019-08-06",
                            username = "ENTER_USERNAME_HERE",
                            password = "ENTER_PASSWORD_HERE",
                            rep = TRUE,
                            rep.name = "grp_hh")

In the command above, it is asking for the main form to be retrieved along with the repeat component with a name grp_hh. The resulting object is a list of two data.frames. The first is the data.frame for the main form and the second is the data.frame for the repeat component named grp_hh.

In the case of the Myanmar MCCT Programme Evaluation, there are multiple repeats within the main form (baseline_mcct_final), the anthropometric form (baseline_mcct_anthro_final) and the village profile form (baseline_villprofile_final). To retrieve all these forms including their respective repeats, the following commands can be used:

## List the various repeat names in baseline_mcct_final
rep.names <- c("grp_hh",              ## HH members dataset
               "support_gov_rep",     ## HH dataset
               "support_ngo_rep",     ## HH dataset
               "support_cso_rep",     ## HH dateset
               "support_ro_rep",      ## HH dataset
               "support_ho_rep",      ## HH dataset
               "support_orgoth_rep",  ## HH dataset
               "child_vc_rep",        ## Vaccination and child illness dataset
               "grp_q2_5_to_q2_7",    ## IYCF dataset
               "ancnow_rep",          ## ANC dataset
               "ancpast_rep")         ## ANC dataset

## Retrieve data and the repeats
surveyData <- get_mcct_data(id = "baseline_mcct_final",
                            start = "2019-08-06",
                            username = "ENTER_USERNAME_HERE",
                            password = "ENTER_PASSWORD_HERE",
                            filename = "main_form",
                            rep = TRUE,
                            rep.name = rep.names)

## List the various repeat names in baseline_mcct_anthro_final
rep.names <- c("grp_family",
               "grp_hh",
               "childanthro_rep",
               "mom_anthro_rep")

## Retrieve anthro datasets and the repeats
anthroData <- get_mcct_data(id = "baseline_mcct_anthro_final",
                            start = "2019-08-06",
                            username = "ENTER_USERNAME_HERE",
                            password = "ENTER_PASSWORD_HERE",
                            filename = "anthro_form",
                            rep = TRUE,
                            rep.name = rep.names)

## List the various repeat names in baseline_villprofile_final
rep.names <- c("respondent_rpt",
               "cbo_yes_grp",
               "credit_rep")

## Retrieve village profile data and the repeats
villageData <- get_mcct_data(id = "baseline_villprofile_final",
                             start = "2019-08-06",
                             username = "ENTER_USERNAME_HERE",
                             password = "ENTER_PASSWORD_HERE",
                             filename = "village_form",
                             rep = TRUE,
                             rep.name = rep.names)

Data cleaning functions

There are two data cleaning functions included in myanmarMCCTdata. One of the functions (clean_hh()) applies cleaning algorithm to the household data which is mainly assignment of unique household identifiers to each row of data and correction of assignment into the three area stratifications. The second (clean_anthro()) applies cleaning algorithm to the anthropometric data which again is mainly assignment of unique household identifiers and correctio of assignment into the three area stratification. These functions are applied to raw household data (hh dataset in the package) and raw anthropometric data pulled from the ONA server.

Data manipulation and structuring functions

Within this family of functions, a further dichotomisation can be created.

One set of functions creates indicator-specific raw datasets that contain the necessary variables to report the indicator/s of interest. This set of function are all prefixed with the verb create_ usually followed by the name of the indicator. These functions are:

• create_mcct() - create raw MCCT indicators variables dataset for each eligible individual

• create_mcct_household() - create raw MCCT indicators variables dataset for each household with at least one eligible household member

• create_canthro() - create raw child anthropometric variables dataset

• create_manthro() - create raw maternal anthropometric variables dataset

• create_iycf() - create raw IYCF indicators variables dataset

• create_anc() - create raw ANC indicators variables dataset

• create_chealth() - create raw child health indicators variables dataset

• create_ppi() - create raw PPI indicators variables dataset

The second set of functions within this family of functions includes utility functions used in the data analysis functions and performs manipulation and structuring of raw datasets in preparation for data analysis. These function perform splitting of datasets in to quintiles (split_to_quintiles()) and calculation of sample sizes per indicator (get_n(), get_n_area() and get_n_wealth()).

Data recoding functions

Once raw data is cleaned and re-structured, it is now ready for recoding. Recoding functions transforms the raw datasets variables into indicator-specific variables based on standard indicator definitions. These functions are all prefixed by the verb recode_ followed by either the name of the variable or the name of the indicator it is recoding. For indicators with complex definitions requiring multi-variable recoding, multiple recode functions are available for each layer or step of the indicator definitions.

Data analysis functions

Recoded indicator datasets can now be used for data analysis. There are two main data analysis functions that apply the complex sample analysis required for this type of study. These functions are:

1. create_weighted_table() - performs Taylor linearised deviation to the recoded indicator dataset and then returns a table of indicator results with sample sizes, indicator estimate, 95% confidence interval and standard error for each of the area stratification (urban, rural, hard-to-reach and total) and each of the wealth quintiles.

2. create_weighted_anthro() - performs Taylor linearised deviation to the recoded indicator dataset and then returns a table of indicator results with sample sizes, indicator estimate, 95% confidence interval and standard error for each of the area stratification (urban, rural, hard-to-reach and total).

Datasets

The myanmarMCCTdata package also comes with various datasets produced by the study. These datasets can be grouped into two sets.

One set of data are the raw datasets retrieved from the ONA server. There are 10 of these datasets included in the package and can be called in R via their data object names. See Reference for the object names and descriptions of these datasets.

The second set of data are the results datasets produced following data analysis. There are 22 of these results datasets included in the package and can be called in R via their data object names. See Reference for the object names and descriptions of these results datasets.

validmeasures/myanmarMCCTdata documentation built on March 29, 2020, 10:27 p.m.