Self-Perceived Loneliness and Depression During the COVID-19 Pandemic: a Two-Wave Replication Study

<p>This is the README file for the scripts of the preprint "Self-Perceived</p> <p>Loneliness and Depression During the COVID-19 Pandemic: a Two-Wave Replication</p> <p>Study" by Carollo et al. (2022)</p> <p><br></p> <p>Access the pre-print here:  <a href="https://eur01.safelinks.protection.outlook.com/?url=https://ucl.scienceopen.com/document/read?vid=0769d88b-e572-48eb-9a71-23ea1d32cecf&data=05|01|j.houghton@ucl.ac.uk|734ebb285cd44530ba8b08da5a92b1ad|1faf88fea9984c5b93c9210a11d9a5c2|0|0|637921883339567203|Unknown|TWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0=|3000|||&sdata=tewZkbNg4VQpgzQfijl+aX6EUqI1dF+M0GjQ4qRKASQ=&reserved=0" target="_blank">https://ucl.scienceopen.com/document/read?vid=0769d88b-e572-48eb-9a71-23ea1d32cecf</a>  </p> <p><br></p> <p>Abstract: Background: The global COVID-19 pandemic has forced countries to</p> <p>impose strict lockdown restrictions and mandatory stay-at-home orders with</p> <p>varying impacts on individual’s health. Combining a data-driven machine learning</p> <p>paradigm and a statistical approach, our previous paper documented a U-shaped</p> <p>pattern in levels of self-perceived loneliness in both the UK and Greek</p> <p>populations during the first lockdown (17 April to 17 July 2020). The current</p> <p>paper aimed to test the robustness of these results by focusing on data</p> <p>from the first and second lockdown waves in the UK. Methods: We tested a) the</p> <p>impact of the chosen model on the identification of the most time-sensitive</p> <p>variable in the period spent in lockdown. Two new machine learning</p> <p>models - namely, support vector regressor (SVR) and multiple linear regressor</p> <p>(MLR) were adopted to identify the most time-sensitive variable in the UK</p> <p>dataset from wave 1 (n = 435). In the second part of the study, we tested</p> <p>b) whether the pattern of self-perceived loneliness found in the first UK</p> <p>national lockdown was generalizable to the second wave of UK lockdown</p> <p>(17 October 2020 to 31 January 2021). To do so, data from wave 2 of the UK</p> <p>lockdown (n = 263) was used to conduct a graphical and statistical inspection</p> <p>of the week-by-week distribution of self-perceived loneliness scores. Results:</p> <p>In both SVR and MLR models, depressive symptoms resulted to be the most</p> <p>time-sensitive variable during the lockdown period. Statistical analysis of</p> <p>depressive symptoms by week of lockdown resulted in a U-shaped pattern</p> <p>between week 3 to 7 of wave 1 of the UK national lockdown. Furthermore,</p> <p>despite the sample size by week in wave 2 was too small for having a meaningful</p> <p>statistical insight, a qualitative and descriptive approach was adopted and</p> <p>a graphical U-shaped distribution between week 3 and 9 of lockdown was</p> <p>observed. Conclusions: Consistent with past studies, study findings suggest</p> <p>that self-perceived loneliness and depressive symptoms may be two of the</p> <p>most relevant symptoms to address when imposing lockdown restrictions.</p> <p><br></p> <p>In particular, the folder includes the scripts for the pre-processing,</p> <p>training, and post-processing phases of the research.</p> <p><br></p> <p>==== PRE-PROCESSING WAVE 1 DATASET ====</p> <p>- "01_preprocessingWave1.py": this file include the pre-processing of the</p> <p>variables of interest for wave 1 data;</p> <p>- "02_participantsexcludedWave1.py": this file include the script adopted to</p> <p>implement the exclusion criteria of the study for wave 1 data;</p> <p>- "03_countryselectionWave1.py": this file include the script to select the UK</p> <p>dataset for wave 1.</p> <p><br></p> <p>==== PRE-PROCESSING WAVE 2 DATASET ====</p> <p>- "04_preprocessingWave1.py": this file include the pre-processing of the</p> <p>variables of interest for wave 2 data;</p> <p>- "05_participantsexcludedWave1.py": this file include the script adopted to</p> <p>implement the exclusion criteria of the study for wave 2 data;</p> <p>- "06_countryselectionWave1.py": this file include the script to select the UK</p> <p>dataset for wave 2.</p> <p><br></p> <p>==== TRAINING ====</p> <p>- "07_MLR.py": this file includes the script to run the multiple regression</p> <p>model;</p> <p>- "08_SVM.py": this file includes the script to run the support vector regression</p> <p>model.</p> <p><br></p> <p>==== POST-PROCESSING: STATISTICAL ANALYSIS ====</p> <p>- "09_KruskalWallisTests.py": this file includes the script to run the multipair</p> <p>and the pairwise Kruskal-Wallis tests.</p> <p><br></p>