Home

Menu

Loading wiki pages...

View
Wiki Version:
<p>This project contains the submitted manuscript draft, main and supplemental analysis scripts, deidentified and aggregated data (compliant with the research project team's policy on sharing data).</p> <p><strong>Instructions for reproducing our reported results:</strong> Researchers interested in reproducing our analyses in R should do the following (note that reliability analyses cannot be reproduced with the shared data). 1. download <a href="https://osf.io/tp6c9/" rel="nofollow">the zipped R project for reproducing our results</a>. 2. unzip this directory 3. open the R project in R Studio (ends in .Rproj and looks like a box) 4. Open the script 'PowersMoshontzHoyle_reproduceanalyses.R' (listed in "Files"), install any required packages as prompted by R Studio, and read the notes at the top for instructions on reproducing all or parts of our results.</p> <p><strong>More on the data we are sharing:</strong> We do not have permission to share the raw data that we used, or some forms of aggregate data (e.g., data with information denoting campus). Some of our analyses therefore cannot be reproduced by people who don't have the raw data. To make our analyses as reproducible and transparent as possible given our data sharing limitations, we share: all analysis scripts, all analysis outputs (as .html Markdown documents), a form of the data that we can share, and a script that uses this data to reproduce all possible analyses.</p> <p><strong>More on the data scripts and output:</strong> Our analyses were conducted in R. The code is commented and comments were written with the intention of being understandable even to people that do not know the programming language. The analysis scripts cannot be run in their entirety without the raw data, but the .html output files shows both code chunks and outputs and can be viewed in any web browser.</p> <p><strong>More on the BCSCS validation analyses:</strong> We report some validity analyses related to our measure of trait self-control. These analyses can be reproduced using steps similar to reproducing our main analyses (<a href="http://osf,io/2sc4z" rel="nofollow">see this zipped project folder</a>). </p> <p><strong>More information about how we did each step of analysis:</strong></p> <p><strong><em>Data Wrangling</em></strong>: We received cleaned data from the project coordinators as a SAS dataset. Using R, we imported these data, created summary scores for our composite measures (see <a href="https://osf.io/rw874/" rel="nofollow">.Rmd script</a> and <a href="https://osf.io/9wd8c/" rel="nofollow">output</a>).</p> <p><em>Note:</em> In the first version of this manuscript (<a href="https://osf.io/download/5caba43366464b0017a21794/?version=2&displayName=PowersMoshontzHoyle-2019-05-01T13:28:09.713Z.pdf" rel="nofollow">described in Version 2 of our preprint</a>) our measure of affect regulation was oddly computed: the project coordinators provided us BriefCOPE subscale scores rather than raw scores, so our adaptive and maladaptive measures were computed from subscales rather than the items individually. Subsequently we requested the raw items and conducted analyses using these (<a href="https://osf.io/download/5caba43366464b0017a21794/?version=3&displayName=PowersMoshontzHoyle-2019-08-12T01:48:08.338Z.pdf" rel="nofollow">described in Version 3 of our preprint</a>). As a result, we import two different raw datasets and then merge them.</p> <p><strong><em>Computing Reliabilities and Descriptives</em></strong>: Project staff prepared demographic summaries based on a list of included study IDs. We did not have access to demographic information about our participants. We computed variable descriptives and reliabilities (alpha and omega) in R with the psych package (see <a href="https://osf.io/rw874/" rel="nofollow">.Rmd script</a> and <a href="https://osf.io/9wd8c/" rel="nofollow">output</a>). We also produced several tables used in the manuscript and summaries that we used while writing and editing the manuscript. When possible, descriptives use the data set we share rather than the raw data. We computed a third measure of reliability, the H index (Hancock & Mueller, 2001), by taking factor scores (see <a href="https://osf.io/k6npm/" rel="nofollow">the R script ana_factorscores</a>) and entering them into <a href="http://osf.io/7a3cb" rel="nofollow">a spreadsheet</a> that was obtained from the website of <a href="http://drjosephhammer.com/research/bifactor-analysis-resources/" rel="nofollow">Dr. Joseph Hammer.</a></p> <p><strong>Normality and t-tests comparing excluded/included and t-tests comparing missing/nonmissing</strong>: We conducted some analyses in preparation for SEM analyses (see <a href="https://osf.io/rw874/" rel="nofollow">.Rmd script</a> and <a href="https://osf.io/9wd8c/" rel="nofollow">output</a>). We examined normality, looked at t-tests to understand how people who were included were different from those who were excluded, and conducted t-tests to compare people in our sample who were missing on the DV to those who were not.</p> <p><strong>SEM analyses</strong>: SEM analyses were conducted in R. In addition to analyses that used our focal independent variable, trait self-control (see <a href="https://osf.io/596qx/" rel="nofollow">.Rmd script</a> and <a href="https://osf.io/pcr5b/" rel="nofollow">output</a>), we also conducted analyses that use conscientiousness in place of trait self-control (as described in the manuscript; see <a href="https://osf.io/rs6kj/" rel="nofollow">.Rmd script</a> and <a href="https://osf.io/2r9mc/" rel="nofollow">output</a>).</p>
OSF does not support the use of Internet Explorer. For optimal performance, please switch to another browser.
Accept
This website relies on cookies to help provide a better user experience. By clicking Accept or continuing to use the site, you agree. For more information, see our Privacy Policy and information on cookie use.
Accept
×

Start managing your projects on the OSF today.

Free and easy to use, the Open Science Framework supports the entire research lifecycle: planning, execution, reporting, archiving, and discovery.