Once the search string was finalized, the project moved towards searching and exporting the full results from each database chosen for the review. The project incorporates a wide range of academic databases (see Table 1, below). In total, to date we have utilised 20 sources. The majority were databases, and we also searched proprietary journal packages, institutional repositories and one search engine (Google Scholar). To ensure reliability and minimise missing information, these electronic sources have been complemented by hand searches of AISHE, the key journal in the Irish teaching and learning field and a citation search of the one existing review of Irish T&L research (Hegarty 2008).
One of the challenges when crafting the search was finding a specific string that could be used across all the sources as not all had the same search functionality. To ensure consistency the team crafted one ‘long and complex’ string for databases with that could facilitate this kind of search, and a backup ‘short and simple’ string for those with more basic functionality. The practice of modifying the search string is not unusual in systematic reviews.
The results gathered in total amounted to 9133 items. After retrieving the results, the next step required us to export these records into the selected reference manager (in this case, Endnote desktop was selected as the most appropriate). After both an automatic and manual removal of duplicate entries within the reference manager, the records have been reduced to 6797 entries.
The core information that is required for our inclusion/exclusion screening process is the reference, abstract, and author address. However, as can be seen from Table 1, not all databases allowed us to export this data to Endnote. While databases such as Web of Science or Scopus have a range of export options allowing the export of precise information, other databases such as Springer, Taylor & Francis, ProQuest, do not allow the automatic export of author address. Google Scholar only allows the export of individual references and does not facilitate the export of either address or abstract. Currently all the incomplete Endnote entries are being supplemented by manual copying of this missing information.
Alongside the export process, the project steering group have been working together screening a sample of records (with complete information) in order to pilot the inclusion/exclusion criteria. As such, the project is poised to move on to the next screening stage of the systematic review.
*Many thanks to Peter Hickey (UCD Library and Project Group) for his very useful feedback on an earlier version of this blog post.
One of the challenges when crafting the search was finding a specific string that could be used across all the sources as not all had the same search functionality. To ensure consistency the team crafted one ‘long and complex’ string for databases with that could facilitate this kind of search, and a backup ‘short and simple’ string for those with more basic functionality. The practice of modifying the search string is not unusual in systematic reviews.
The results gathered in total amounted to 9133 items. After retrieving the results, the next step required us to export these records into the selected reference manager (in this case, Endnote desktop was selected as the most appropriate). After both an automatic and manual removal of duplicate entries within the reference manager, the records have been reduced to 6797 entries.
The core information that is required for our inclusion/exclusion screening process is the reference, abstract, and author address. However, as can be seen from Table 1, not all databases allowed us to export this data to Endnote. While databases such as Web of Science or Scopus have a range of export options allowing the export of precise information, other databases such as Springer, Taylor & Francis, ProQuest, do not allow the automatic export of author address. Google Scholar only allows the export of individual references and does not facilitate the export of either address or abstract. Currently all the incomplete Endnote entries are being supplemented by manual copying of this missing information.
Alongside the export process, the project steering group have been working together screening a sample of records (with complete information) in order to pilot the inclusion/exclusion criteria. As such, the project is poised to move on to the next screening stage of the systematic review.
*Many thanks to Peter Hickey (UCD Library and Project Group) for his very useful feedback on an earlier version of this blog post.