Welcome to the ultimate 'data buffet' – MetaReviewer style. Our datasets are all you can analyze, so grab your digital fork and dive in. In this post, we explain how to download all the data your team works hard to collect!
What is downloadable?
Users have access to three downloadable files: citations portal data, studies portal data, and screening/coding form response data.
Accessing data from the citations and studies portals doesn’t require any navigation outside of going to those portals. Once there – click the download button.
Accessing collected information from a screening or coding form requires a few more steps. First, navigate to the project management page. Second, click on “Manage coding forms and view responses” in the panel on the right side of the page. Third, navigate to the screening or coding form row that contains data to export. Fourth, click on the icon that is second from the left on the right-hand side of the screen – if you hover over it, the caption says “View responses.” Fifth and finally, once there, click on the download button.
How are they structured?
Starting with the ‘easier’ downloads, data from the citations and studies portals are WYSIWYG – what you see is what you get. Each row in the portals translates to a row in the downloaded dataset.
Here’s a screenshot of an ongoing project’s Studies Portal page.
And here’s the associated spreadsheet that was exported.
Responses from each screening or coding form are captured in separate, downloadable datasets. As a result, should you use MetaReviewer to conduct full-text screening and full-text data extraction, then you will have two different datasets available to download (one for full-text screening and one for data extraction).
The resulting dataset exported from full-text screening is relatively straightforward. Our suggested full-text screening form template creates a screening form with one page. This means that each study will contain one set of responses determining its eligibility. When you export the dataset associated with the full-text screening coding form, each row will represent the single study where users enter eligibility information.
On the other hand, the resulting dataset from a data extraction coding form is less straightforward, at least, in terms of how the data are combined. The layout is simple enough – every effect size that a user enters results in another row in the exported dataset. Every column in the dataset is a field that a user creates, for instance when using one of our coding form templates.
The complexity lies in the connection between effect sizes and the description of those effects. Let’s use an example that might often occur in evidence synthesis practice: a single intervention study reports effect size information for multiple outcome measures, say reading and writing test scores. In the Measures page, a coder creates two columns representing each outcome.
Then, on the Effect Size page, a coder creates two rows representing the effect size for each of the outcomes.
Now when a user exports the coding form dataset, MetaReviewer will automatically connect (i.e., link) the effect size data to the measures data. That is, effect size row 1 will be linked to the Reading measure and all the information that a user stores about the Reading measure will be exported. Same for the Writing measure; both can be seen in the spreadsheet screenshot below.
This information is linked to the appropriate effect size row, along with the remaining connected information from the Study, Sample, Condition, and Quality pages.
When are they useful?
We use exported datasets for two primary reasons: project management and analysis.
Although we are continually adding and updating project management features, like study tagging, we know users need more control over progress and tracking. If this speaks to you, then we suggest downloading the Studies Portal page and using it as an up-to-date progress tracker. In our own practice, we often create simple Excel pivot tables that provide status summaries on all sorts of screening and coding parameters. Eventually we will automate these summary reports in MetaReviewer, but until then, users will find data downloads helpful.
One additional application for exported datasets is to conduct the quantitative synthesis or meta-analysis. Because the data extraction coding form dataset represents each effect size and its associated information, users can immediately begin formatting, processing, and analyzing the dataset. In fact, we have used the resulting data extraction coding form dataset to complete three recent meta-analyses.
Now that you know how to get the data you worked hard to collect, download it and start your next analysis. Bon Appétit, MetaReviewer enthusiasts!
