---
title: "Usage"
output: rmarkdown::html_vignette
vignette: >
%\VignetteIndexEntry{Usage}
%\VignetteEngine{knitr::rmarkdown}
%\VignetteEncoding{UTF-8}
---
```{r, include = FALSE}
knitr::opts_chunk$set(
collapse = TRUE,
comment = "#>"
)
```
# Tutorial package calidad
The package aims to implement in a simple way the methodologies of [INE Chile](https://www.ine.cl/docs/default-source/documentos-de-trabajo/20200318-lineamientos-medidas-de-precisi%C3%B3n.pdf?sfvrsn=f1ab2dbe_4), [ECLAC 2020](https://repositorio.cepal.org/bitstream/handle/11362/45681/S2000293_es.pdf?sequence=4&isAllowed=y) and [ECLAC 2023](https://repositorio.cepal.org/server/api/core/bitstreams/f04569e6-4f38-42e7-a32b-e0b298e0ab9c/content) for the quality assessment of estimates from household surveys and [INE Chile Economics Surveys 2024](https://www.ine.gob.cl/docs/default-source/buenas-practicas/directrices-metodologicas/estandares/documentos/est%C3%A1ndar-evaluaci%C3%B3n-de-calidad-de-estimaciones-econ%C3%B3micas.pdf?sfvrsn=201fbeb9_2) for the quality assessment of estimates from economics surveys.
This tutorial shows the basic use of the package and includes the main functions to create the necessary inputs to implement the quality standards.
## Data edition
We will use three datasets:
- Encuesta Nacional de Empleo (efm 2020)
- VIII Encuesta de Presupuestos Familiares
- VII Encuesta longitudinal de Empresas (ELE7)
Those datasets are loaded into the package and they can be used when the package is loaded in the session [^haven]. The data edition in the case of ENE has the purpuse of creating some subpopulations (work force, unemployed and unemployed).
[^haven]: The data contained within the package has some editions. It is important to note that haven::labelled objects may have some collision with quality functions. If you want to import a dta file, all variables that are of type haven::labelled must be converted to numeric or character.
```{r, message=FALSE, warning=FALSE}
library(survey)
library(calidad)
library(dplyr)
ene <- ene %>%
mutate(fdt = if_else(cae_especifico >= 1 & cae_especifico <= 9, 1, 0), # labour force
ocupado = if_else(cae_especifico >= 1 & cae_especifico <= 7, 1, 0), # employed
desocupado = if_else(cae_especifico >= 8 & cae_especifico <= 9, 1, 0),
hombre = if_else(sexo == 1, 1, 0),
mujer = if_else(sexo == 2, 1, 0)) # unemployed
# One row per household
epf <- epf_personas %>%
group_by(folio) %>%
slice(1) %>%
ungroup()
```
## Sample design
Before starting to use the package, it is necessary to declare the sample design of the survey, for which we use the `survey` package. The primary sample unit, the stratum and weights must be declared. It is also possible to use a design with only weights, nevertheless in that case the variance will be estimated under simple random sampling assumption. In this case we will declare a complex design for the two surveys (EPF and ENE). Additionally, it may be useful to declare an option for strata that only have one PSU.
```{r, results='hide'}
# Store original options
old_options <- options()
```
```{r}
# Complex sample design for ENE
dc_ene <- svydesign(ids = ~conglomerado , strata = ~estrato_unico, data = ene, weights = ~fact_cal)
# Complex sample design for EPF
dc_epf <- svydesign(ids = ~varunit, strata = ~varstrat, data = epf, weights = ~fe)
options(survey.lonely.psu = "certainty")
```
For ELE survey, we also must to declare the fpc (finite population correction)
```{r, eval=FALSE}
# Complex sample design for ELE
dc_ele <- svydesign(ids = ~rol_ficticio, weights = ~fe_transversal, strata = ~estrato, fpc = ~pob, data = ELE7)
options(survey.lonely.psu = 'remove')
```
## Inputs creation
### National Labour Survey (part 1)
To assess the quality of an estimate, the [INE methodology](https://www.ine.cl/docs/default-source/documentos-de-trabajo/20200318-lineamientos-medidas-de-precisi%C3%B3n.pdf?sfvrsn=f1ab2dbe_4) establishes differentiated criteria for estimates of proportion or ratio between 0 and 1, on the one hand, and estimates of mean, size and total, on the other. In the case of proportion estimation, it is necessary to have the sample size, the degrees of freedom and the standard error. The other estimates require the sample size, the degrees of freedom, and the coefficient of variation.
The package includes separate functions to create the inputs for estimates of **mean, proportion, totals and size**. The following example shows how the proportion and size functions are used.
```{r}
insumos_prop <- create_prop(var = "desocupado", domains = "sexo", subpop = "fdt", design = dc_ene) # proportion of unemployed people
insumos_total <- create_size(var = "desocupado", domains = "sexo", subpop = "fdt", design = dc_ene) # number of unemployed people
```
- `var`: variable to be estimated. Must be a dummy variable
- `domains`: required domains.
- `subpop`: reference subpopulation. It is optional and works as a filter (must be a dummy variable)
- `design`: sample design
The function returns all the neccesary inputs to implement the standard
```{r}
insumos_total
```
To get more domains, we can use the "+" symbol as follows:
```{r}
desagregar <- create_prop(var = "desocupado", domains = "sexo+region", subpop = "fdt", design = dc_ene)
```
A useful parameter is `eclac_input`. It allows to return the ECLAC inputs. By default this parameter is `FALSE` and with the option `TRUE` we can activate it.
```{r, eval=F}
eclac_inputs <- create_prop(var = "desocupado", domains = "sexo+region", subpop = "fdt", design = dc_ene, eclac_input = TRUE)
```
In some cases, it may be of interest to assess the estimate of a ratio. The `create_prop` function allows ratio estimation by explicitly specifying the numerator and the denominator. To do this, the function includes the denominator parameter, where the variable to be used in the denominator of the ratio must be provided.
For ECLAC standards, it is necessary to specify the version of the methodology using the scheme parameter, since ratio estimation is only supported under the ECLAC 2023 standard (`eclac_2023`). The default option is `eclac_2020`, which does not include support for ratio estimation. An example using `eclac_2023` is shown below:
```{r, eval=FALSE, warning=FALSE}
create_prop(var = "mujer", denominator = "hombre", domains = "ocupado", design = dc_ene,
eclac_input = TRUE, scheme = 'eclac_2023')
```
### Household Budget Survey (part 2)
We may also be interested in assessing the estimate of a sum. For example, the sum of all the income of the EPF at the geographical area level (Gran Santiago and other regional capitals). For this, there is the `create_total` function. This function receives a continuous variable such as hours, expense, or income and generates totals at the requested level. The ending "with" of the function alludes to the fact that a continuous variable is being used.
```{r, eval=T, warning=FALSE}
insumos_suma <- create_total(var = "gastot_hd", domains = "zona", design = dc_epf)
```
If we want to assess the estimate of a mean, we have the function `create_mean`. In this case, we will calculate the average expenditure of households, according to geographical area.
```{r}
insumos_media <- create_mean(var = "gastot_hd", domains = "zona", design = dc_epf)
```
The default usage is not to disaggregate, in which case the functions should be used as follows:
```{r}
# ENE dataset
insumos_prop_nacional <- create_prop("desocupado", subpop = "fdt", design = dc_ene)
insumos_total_nacional <- create_total("desocupado", subpop = "fdt", design = dc_ene)
# EPF dataset
insumos_suma_nacional <- create_total("gastot_hd", design = dc_epf)
insumos_media_nacional <- create_mean("gastot_hd", design = dc_epf)
```
The create functions also include a `ci` parameter, which allows returning the confidence interval. By default, this parameter is set to `FALSE.` For proportions, the logistic confidence interval is available. To use it, you need to set `ci_logit = TRUE`, as shown below:
```{r}
# ENE dataset
prop_nacional_ci <- create_prop("desocupado", subpop = "fdt", design = dc_ene, ci = TRUE)
prop_nacional_ci_logit <- create_prop("desocupado", subpop = "fdt", design = dc_ene, ci_logit = TRUE)
```
### Longitudinal Enterprise Survey (part 3)
In the case of economic surveys, the estimates are carried out in a similar way. For example, in the Longitudinal Enterprise Survey (ELE), wage productivity is estimated as the ratio between value added and the total gross remuneration of personnel hired by the company.
```{r, eval=FALSE}
prod_salarial <- create_prop('VA_2022f', denominator = 'REMP_TOTAL', domains = 'cod_actividad+cod_tamano', design = dc_ele)
```
## Assessment
Once the inputs have been generated, we can proceed with the assessment using the `assess` function. By default, the INE Chile criteria are applied (`scheme = 'chile'`). However, if we want to evaluate using the ECLAC or Chile Economics criteria, we need to specify `scheme = 'eclac_2020'`, `scheme = 'eclac_2023'`, or `scheme = 'chile_economics'`.
For the ECLAC 2023 standard (`eclac_2023`), we can use the parameters `domain_info` and `low_df_justified` to indicate whether the estimation corresponds to a planned study domain and whether low degrees of freedom are justified. Both parameters are set to `FALSE` by default.
```{r, eval=F, warning=FALSE}
# INE Chile
evaluacion_prop <- assess(insumos_prop)
evaluacion_tot <- assess(insumos_total)
evaluacion_suma <- assess(insumos_suma)
evaluacion_media <- assess(insumos_media)
# ECLAC
evaluacion_cepal_2020 <- assess(eclac_inputs, scheme = 'eclac_2020')
evaluacion_cepal_2023 <- assess(eclac_inputs, scheme = 'eclac_2023', domain_info = TRUE, low_df_justified = TRUE)
```
In the INE Chile Economics standard (`chile_economics`), the `domain_info` parameter is also available, and the `table_n_obj` parameter can be used to specify the target sample size for the domains to be evaluated, using a data frame. This data frame must include the target sample size in a column named `n_obj`. By default, `table_n_obj` is `NULL`, and the sample recovery verification will be skipped.
For the assessment of ratios, we can indicate whether the estimated value falls between 0 and 1 using the parameter `ratio_between_0_1`. By default, it is set to `TRUE`. If a value greater than 1 is detected, the coefficient of variation (CV) will be used instead.
```{r, eval = FALSE}
# INE Economics
## target sample size
table_n_obj <- ELE7_n_obj %>%
dplyr::mutate(cod_actividad = cod_actividad_letra,
cod_tamano = as.character(cod_tamano)) %>%
dplyr::select(-cod_actividad_letra)
eval_ratio <- assess(prod_salarial, scheme = 'chile_economics',
domain_info = TRUE, table_n_obj = table_n_obj, ratio_between_0_1 = FALSE)
```
The output is a `dataframe` that, in addition to containing the information already generated, includes a column that indicates whether the estimate is unreliable, less reliable or reliable.
The function `assess` has a parameter that allows us to know if the table should be published or not. Following the criteria of the standard, if more than 50% of the estimates of a table are not reliable, it should not be published.
```{r eval=F}
# Unemployment by region
desagregar <- create_size(var = "desocupado", domains = "region", subpop = "fdt", design = dc_ene)
# assess output
evaluacion_tot_desagreg <- assess(desagregar, publish = T)
evaluacion_tot_desagreg
```
```{r eval=F}
# Reset original options
options(old_options)
```