theory-tsq.Rmd

# Total survey quality  {#tsq}

:::: { .warn}
<img src="images/warning.svg" class="svg-inline"/> **Work in progress!** Please come back later.  
::::
<!--
As we have seen in Chapter \@ref(tse), low levels of TSE are necessary for a survey estimator to be accurate. While accuracy is necessary for quality, it is not sufficient. Survey data must also respond to user needs in order to yield useful results. Data users often take data accuracy as given, assuming that it has been controlled by data producers. Instead they prioritize properties such as data being rich in detail, easily accessible and well documented. Data that is not sufficiently detailed, difficult to access or hard to interpret may be unfit for use and be perceived to have low quality by data users, even if being accurate.

## Dimensions of survey quality

The concept of *total survey quality* (TSQ) considers the *fitness of use* of a survey estimate. It combines the accuracy of an estimate with non-statistical dimensions oriented towards the data user. The definitions of survey quality used by most NSOs in Europe, North America and Oceania, as well as that of international organisations such as Eurostat, IMF, and OECD explicitly acknowledge the multi-dimensional nature of survey quality. These definitions are referred to as *survey quality frameworks*. There is some (often minor) variation between organisations, but most frameworks include a subset of the quality dimensions summarized in Table \@ref(tab:qd) [@Biemer2010].

```{r, echo = FALSE, warning = FALSE, comment = FALSE, message = FALSE}
library(kableExtra)
options(kableExtra.html.bsTable = T)
library(dplyr)
quality <- tribble(~Dimension, ~Description,
     "Accuracy ", "Total survey error is minimized ",
     "Credibility ", 	"Data are considered trustworthy by the survey community", 
     "Comparability ",	"Demographic, spatial, and temporal comparisons are valid", 
     "Usability/Interpretability", "Documentation is clear and metadata are well-managed", 
     "Relevance ", "Data satisfy users' needs", 
     "Accessibility ",	"Access to the data is user friendly",
     "Timeliness/Punctuality", "Data deliveries adhere to schedules",
     "Completeness ", "Data are rich enough to satisfy the analysis objectives without undue burden on respondents", 
     "Coherence ",	"Estimates from different sources can be reliably combined")

quality %>%
  kable(caption = 'Common Dimensions of a Survey Quality Framework.', label = "qd") %>%
  kable_styling(bootstrap_options = c("striped")) 
```

These frameworks are commonly used as the basis of survey quality reporting. Some of the dimensions are qualitative, making it difficult to generate a single metric to summarize the overall quality of a survey. Survey methodologists have not (yet) put forward a standard quality measure. Instead, evaluations tend to identify the strength and weaknesses of a survey by dimension and assess how well it achieved the goals of each dimension. 

### Accuracy

needs to minimize total survey error

### Timeliness

Another very useful application of survey quality frameworks is as a design principle to maximize the TSQ in a survey.   

@StatCan2019

## Maximising quality
-->