Table 3

Searching and analysis practices reported in order of frequency of use

Searching practices	Searching practices continued	Analysis practices	Analysis practices continued
Search strings or terms listed	Notes more documents returned from some companies than others	Authors stated they kept “relevant” documents	Quantitative techniques used
Searching by topic or keyword	Date range of documents returned listed	Abstracting or summarising	Coders were trained
Number of documents found listed	Searching labelled “exhaustive”	Selection criteria described	Analysis labelled “qualitative”
Describes use of snowballing	Searching by title	Documents ordered chronologically	Analysis labelled “mixed methods”
Searching for people or names	Searching labelled “systematic”	Documents coded or indexed	Authors described considering quality of document information in analysis
Dates of search listed	Searching for projects or accounts	Specific analysis methods named (e.g. case study methods, content analysis)	Analysis labelled “analytical”
Searching for organisations committees or companies	Searching labelled “normative”	More than one coder used	Authors described interpreting the documents against other material
Search terms combined	Physical searching in depositories	Themes identified	Authors described emphasising documents for which a context was available
Searching by dates	Searching by request for production codes	Documents coded inductively	Authors described emphasising repeated rather than single instances
Overall search strategy described	Searching by document type	Analysis labelled “narrative”	Analysis labelled “historical”
Explains why search strategy used	Uses optical character recognition	Analysis labelled “descriptive”
Searching by source, file or location	Searching for meetings or events	Statistical tests used
Searching by Bates numbers	Used keyword variation (e.g. misspellings)	Software used to manage data