`datsearch.Rd`

Function to search among the existing datasets.

`datsearch(pattern, concept=TRUE, matchall=TRUE, fixed=TRUE, pkgdown=FALSE)`

- pattern
character string or vector of strings specifying the terms to search for within the datasets. Can also be left unspecified to start the function in an interactive mode.

- concept
logical indicating whether the search should be confined to the concept terms (

`TRUE`

by default) or whether a full-text search should be conducted.- matchall
logical indicating whether only the datasets matching all terms (if multiple are specified) are returned (

`TRUE`

by default) or whether datasets matching any one of the terms are returned.- fixed
logical indicating whether a term is a string to be matched as is (

`TRUE`

by default). If`FALSE`

, a search term is a regular expression that`grep`

will search for. Only relevant when`concept=FALSE`

(i.e., when doing a full-text search).- pkgdown
logical indicating whether the standard help file or the pkgdown docs (at https://wviechtb.github.io/metadat/) should be shown for a chosen dataset (

`FALSE`

by default).

The function can be used to search all existing datasets in the metadat package based on their concept terms (see below) or based on a full-text search of their corresponding help files.

When running `datsearch()`

without the `pattern`

argument specified, the function starts in an interactive mode and prompts for one or multiple search terms.

Alternatively, one can specify a single search term via the `pattern`

argument or multiple search terms by using a string vector as the `pattern`

or by separating multiple search terms in a single string with ‘,’, ‘;’, or ‘and’.

If `matchall=TRUE`

(the default), only datasets matching all search terms (if multiple are specified) are returned. If `matchall=FALSE`

, datasets matching any one of the search terms are returned.

If a single match is found, the corresponding help file is directly shown. If multiple matches are found, the user is prompted to choose one of the matching datasets of interest.

**Concept Terms**

Each dataset is tagged with one or multiple concept terms that refer to various aspects of a dataset, such as the field/topic of research, the outcome measure used for the analysis, the model(s) used for analyzing the data, and the methods/concepts that can be illustrated with the dataset.

In terms of ‘fields/topics’, the following terms have been used at least once: alternative medicine, attraction, cardiology, climate change, covid-19, criminology, dentistry, ecology, education, engineering, epidemiology, evolution, genetics, human factors, medicine, memory, obstetrics, oncology, persuasion, physics, primary care, psychiatry, psychology, smoking, social work, sociology.

In terms of ‘outcome measures’, the following terms have been used at least once: correlation coefficients, Cronbach's alpha, hazard ratios, incidence rates, raw mean differences, odds ratios, proportions, ratios of means, raw means, risk differences, risk ratios, (semi-)partial correlations, standardized mean changes, standardized mean differences.

In terms of ‘models/methods/concepts’, the following terms have been used at least once: cluster-robust inference, component network meta-analysis, cumulative meta-analysis, diagnostic accuracy studies, dose response models, generalized linear models, longitudinal models, Mantel-Haenszel method, meta-regression, model checks, multilevel models, multivariate models, network meta-analysis, outliers, Peto's method, phylogeny, publication bias, reliability generalization, single-arm studies, spatial correlation.

```
# note: the examples below are not run since they require interactivity
if (FALSE) {
# start the function in the interactive mode
datsearch()
# find all datasets tagged with the concept term 'standardized mean differences'
datsearch("standardized mean differences")
# find all datasets tagged with the concept terms 'odds ratio' and 'multilevel'
datsearch("odds ratio, multilevel")
# do a full-text search for the term 'infarct'
datsearch("infarct", concept=FALSE)
# do a full-text search for 'rma.mv(' (essentially finds all datasets where
# the rma.mv() function was used in the examples section of a help file)
datsearch("rma.mv(", concept=FALSE)
}
```