Stata ADO that matches two columns or two datasets based on similar text patterns

Data in two columns in the same dataset

 matchit varname1 varname2 [, options]

Data in two different datasets (with indexation)

 matchit idmaster txtmaster using filename.dta , idusing(varname) txtusing(varname) [options]

String matching method. Default is bigram. Other typical built-in simfcn are:ngram, ngram_circ, token, soundex and token_soundex.

Specifies similarity score. Default is jaccard. Other built-in options are simple and minsimple.

Weighting transformation. Default is noweights. Other built-in options are simple, log and root.

Specifies the name for the similarity score variable. Default is similscore.

Numeric varname from current file (masterfile).
Needs not to uniquely identify observations from masterfile (although recommended).

String varname from current file (masterfile) which will be matched to txtusing.

Name (and path) of the Stata file to be matched (usingfile).

Numeric varname from usingfile.
Needs not to uniquely identify observations from usingfile (although recommended).

String varname from usingfile which will be matched to txtmaster.

Allows loading weights from a Stata file, instead of computing it from the current dataset (and using dataset, in the case of two-dataset setup). Default is not to load weights.

Outputs time stamps during the execution. To be used for benchmarking purposes.

Controls how often matchit reports back to the output screen.
Only really useful for optimizing indexation by trying different simfcn.
Default is step = 20 (percent).

Lowest similarity scores to be kept in final results. Default is num = .5.

Ignores unsaved data warning.

Reports a preliminary analysis about indexation.
To be used for optimizing indexation by cleaning original data and trying different simfcn.

Generates list of stopwords automatically.
It improves indexation speed but ignores potential matches.

Only valid with stopwordsauto.
It sets the threshold of grams per observation to be included in the stopwords list.
Default is grams-per-observation = .2.

matchit provides a similarity score between two different text strings by performing many different string-based matching techniques. It returns a new numeric variable (similscore) containing the similarity score, which ranges from 0 to 1. A similscore of 1 implies a perfect similarity according to the string matching technique chosen and decreases when the match is less similar. similscore is a relative measure which can (and often do) change depending on the technique chosen. For more information on these techniques refer to Raffo & Lhuillery (2009).

These two variables can be from the same dataset or from two different ones. This latter option makes matchit a convenient tool to join observations when the string variables are not necessarily exactly the same. In other terms, it allows for the dataset currently in memory (called the master dataset) to be matched with filename.dta (called the using dataset) by means of a fuzzy similarity between string variables of each dataset. In this case, matchit returns a new dataset containing five variables: two from the master dataset (idmaster and txtmaster), two from the using dataset (idusing and txtusing) and the already mentioned similarity score (similscore).

matchit is particularly useful in two cases: (1) when the two columns/datasets have different patterns for the same string data (e.g. individual or firm names, addresses, etc.); and, (2) when one of the datasets is considerably large and it was feeded by different sources, making it not uniformly formatted (e.g. names or addresses in different orders). Joining data in cases like these may lead to several false negatives when using merge or similar commands.

matchit is intended for overcoming this kind of problems without engaging into extensive data cleaning or correction efforts. Take, for instance, a case like (1) where one dataset contains first and last names in separated fields, while the other one has just a fullname field. The use of matchit allows to join the two datasets by simply combining the two fields of the first dataset without caring about the order of first and last names or about missing middle names. Similarly, a typical example of (2) is a large dataset containing addresses entered as free-text by different people. Using matchit you can join them with a more standardized source without caring if the zip or state codes were added systematically or not.

Please, note that matchit is case-sensitive. It also takes into account all other symbols (as far as Stata does). While data cleaning is not needed for using matchit, it often implies an improvement of the similarity scores and, in consequence, the overall quality of the matching exercise. However, too much data cleaning might remove relevant information, inducing a negative effect on quality due to false positives.

matchit requires freqindex to be installed when computing weights.