SAD disaster study

This is the dataset associated with the SAD disaster study. Here is the part of the methods as submitted to the Scandinavian Journal of Psychology, where the study is currently under review: Ethics This study was approved by the National Bioethics Committee of Iceland (Approval number: VSN-24-085). Written informed consent was obtained from all participants in the form of information as part of the first page of the online survey. Only participants who acknowledged the information would proceed to the survey questions. All data was collected anonymously. Recruitment Participants for the online survey were recruited via a paid Facebook advertisement from June to September 2024. This advertisement specifically targeted people who lived in specific risk areas in Iceland and, therefore, were living under the risk of experiencing a natural disaster and were likely to have experienced natural disasters in the past. Furthermore, we selected control areas that were in close proximity to these risk areas but were clearly without the specific risk. Additionally, the link to the advertisement was shared to dedicated Facebook groups for members who lived in these areas of interest. Please note the specific areas are not listed here due to the sensitive nature of the data collected in a small community of Iceland. Risk areas were defined as those where natural disasters of significant dimension (i.e., requiring evacuation and/or having caused major destruction of buildings etc.) have happened in the past and can happen again. Psychological Questionnaires The study was fully based on self-report questionnaires delivered via an online survey. The online survey included demographic questions regarding current area of residence and duration of residence in that area to determine natural disaster threat and exposure, gender, age at inclusion (measured as age group), marital status, children (yes/no), educational level, and employment. The online survey also included standardized questionnaires about seasonal symptoms (Seasonal Pattern Assessment Questionnaire, SPAQ; (Rosenthal et al., 1987)), ruminative response style (Ruminative Response Scale, RRS; (Treynor et al., 2003)), anxiety (Generalised Anxiety Disorder Assessment, GAD-7; (Löwe et al., 2008; Spitzer et al., 2006)), perception of risk about imminent danger of natural disaster (Perceived Risk Scale, PRS; (Margarint et al., 2021)), satisfaction with life (Satisfaction with Life Scale, SWLS; (Diener et al., 1985)), depression (Patient Health Questionnaire, PHQ-9; (Kroenke et al., 2001)), eco-anxiety (Hogg-Eco Anxiety Scale, HEAS; (Hogg et al., 2021)), and disaster anxiety (Disaster Anxiety Scale, DAS; (Güzel, 2022)). Additionally, we asked specific questions about the experience of natural disasters, how long ago the natural disaster happened, whether safety was threatened, whether the participants had to evacuate their homes, and the length of evacuation, if applicable. The full set of questions was implemented in Microsoft Forms. Since this survey software does not allow for automatic control of repeated participation in the study, we manually screened the results for repeated participation based on the similarity of demographic information entered. SPAQ, RRS, GAD-7, PHQ-9, HEAS and SWLS have all previously been translated to Icelandic and were used in their Icelandic version. DAS and PRS were used and translated with permission of the authors according to a translation-backtranslation procedure as follows: First, the questionnaires were translated from English to Icelandic independently by two native Icelandic speakers with university level education in psychology. Then, they were back translated to English independently by two additional Icelandic native speakers. The conformity of the backtranslation with the original English version was verified by three expert researchers in psychology, one Icelandic native speaker and two who are highly proficient in English. The SPAQ is a questionnaire used to measure seasonal patterns, including changes in mood and behaviour between seasons. Rosenthal et al. (1987) developed the scale which allows self-report of seasonal changes in appetite, weight, sleep, energy, socialising behaviour, and mood on a 5-point Likert scale with answer options ranging from “no change” (0 points) to “extremely marked change” (4 points). The total score is known as the Global Seasonality Score (GSS), where points of 11 or higher are categorized as high seasonality (Magnusson, A., 1996; Rosenthal et al., 1987). The questionnaire was validated in Iceland (Magnusson, A., 1996; Magnússon & Stefánsson, 1993), with a resulting sensitivity of 94%, specificity of 73% and positive predictive value of 45% for seasonal affective disorder and subsyndromal seasonal affective disorder, combined. The SWLS is used to measure global satisfaction and is a 5-item questionnaire (Diener et al., 1985). It uses a 7-point Likert scale, where answering options range from “strongly disagree” (1 point) to “strongly agree” (7 points). The scale has shown a good test-retest correlation coefficient (α = 0.50 - 0.82 depending on time between test and re-test) and strong internal consistency (α = 0.79 – 0.89) (Diener et al., 1985; Pavot & Diener, 1993). The PHQ-9 contains 9 items focused on the presence of depressive symptoms according to formal diagnostic criteria for the last two weeks and one additional item that focuses on level of impairment due to the symptoms (Kroenke et al., 2001). The questions are scored on a scale from “not at all” (0 points) to “nearly every day” (3 points). The threshold of the total score is five for mild depressive symptoms, 10 for moderate, 15 for moderate severe, and 20 for severe symptoms of depression (Kroenke et al., 2001). Internal consistency of the PHQ-9 is excellent (α = 0.83 – 0.89) as well as re-test reliability (α = 0.84) (Cameron et al., 2008; Kroenke et al., 2001). The GAD-7 is a 7-item questionnaire created to identify generalized anxiety disorder, its characteristics, and symptoms severity for the last two weeks (Löwe et al., 2008; Spitzer et al., 2006). The questions are scored with answer options from “not at all” (0 points) to “nearly every day” (3 points). The cut-off points are five for mild anxiety symptoms, 10 for moderate, and 15 for severe (Spitzer et al., 2006). It has a sensitivity of 89% and a specificity of 82% (Spitzer et al., 2006), internal consistency of α = .89 (Löwe et al., 2008) and test-retest-reliability ICC = .83 (Spitzer et al., 2006). The RRS is used to measure rumination tendencies, and the short form contains 10 items, five that measure depressive brooding and five that measure depressive reflection (Treynor et al., 2003). The RRS items describe responses to depressed mood that are self-focused, symptom-focused, and focused on the possible consequences and causes of their mood (Treynor et al., 2003). The items in the questionnaire are scored on a 4-point scale, ranging from “never or almost never” (1 point) to “always or almost always” (4 points). In the current study we used the sum score of the five questions that measure depressive brooding (RRSb) that represents a passive and abstract way of thinking about ones feelings and symptoms (Treynor et al., 2003). The scale has shown acceptable convergent and predictive validity (Nolen-Hoeksema et al., 1994). The Icelandic translation has acceptable psychometric properties (Hjartarson et al., 2021). The HEAS-13 is used to measure anxiety about environmental conditions, perception of one’s own negative impact on the planet within 13 items referring to the last two weeks (Hogg et al., 2021). Each item has four response options ranging from “not at all” (0 points) to “nearly every day” (3 points) (Hogg et al., 2021). The HEAS-13 has shown good internal reliability (α = 0.92) (Hogg et al., 2021). The DAS is used to measure the presence and level of disaster anxiety in a 6-item questionnaire scored by six points (Güzel, 2022). In the original version of the DAS, response options are zero for “no anxiety”, one for “almost no anxiety”, two for “mild anxiety”, three for “moderate anxiety”, four for “severe anxiety” and five for “extreme anxiety”. When translating the questionnaire to Icelandic, we changed the response options to match better the questions that ask about the frequency of occurrence of different anxiety symptoms. The new answer options are “never” (0 points), “almost never” (1 point), “very rarely” (2 points), “rarely” (3 points), “often” (4 points), “Very often” (5 points). We also modified the instructions such that participants were asked the following: “When responding to the statements below, please consider earthquakes, avalanches, landslides, or volcanic eruptions”. The scale has shown good internal reliability (α = 0.96) and test-retest reliability (0.96). The PRS is a 19-item questionnaire designed to measure risk perception. For this study, seven questions were selected and adapted with permission from the authors to specifically assess risk perception for natural hazards (Margarint et al., 2021). The selected questions included items such as: “Considering a set of natural hazards, which of these events could be a threat/danger to your community?”; “Considering a set of natural hazards, which of these events could be a threat/danger to you personally?”; and “Considering a set of natural hazards, what is the probability that these events could happen in the place where you live or nearby?”. These three questions were answered each for the four selected natural hazards relevant to Iceland, specifically avalanches, earthquakes, mudslides, and volcanic eruptions. The total of these three questions for four types of hazards was used to form a risk perception score by summing up the answers which were scored on 5-point Likert scales ranging from “very low probability” (1) to “very high probability” (5). Another item, “Do you think that these events could be a more frequent threat/danger to the next generations?” was included with a dichotomous response option (“yes” or “no”). To explore potential contributing factors, we also incorporated the question, “Which factors do you think might exacerbate the negative consequences of natural hazards?” This was again rated on a 5-point Likert scale from “very low probability” (1) to “very high probability” (5) and included subcategories such as climate change, deforestation, lack of protective structural devices, poor maintenance of such devices, and the construction of buildings in high-risk areas. Two additional questions focused on risk preparedness: “Considering a set of natural hazards, how much do you feel prepared to cope with these events?” and “Considering a set of natural hazards, how prepared is your community to cope with these events?” These were also scored on 5-point Likert scales ranging from “very badly prepared” (1) to “very well prepared” (5) and were answered again for the 4 relevant natural hazards. The scores on these two questions were summed up across the four types of hazards and formed the risk preparedness score. For both the risk perception score and risk preparedness score we calculated a seasonal and non-seasonal subscore by splitting the answers into those relating to seasonal disasters (avalanches and mudslides) and non-seasonal disasters (volcanic eruptions and earthquakes). Disaster threat exposure We created a factor of general disaster exposure by dividing participants into a disaster group and a control group. Inclusion in the disaster group was based on at least one of the following criteria: a. The participant reported that natural disasters have occurred in their current residence. b. The participant indicated to live in a town where natural disasters have occurred in the past. c. The participant reported that their safety had been threatened by a mudslide, avalanche, earthquake, or volcanic eruption. d. The participant reported having evacuated their home due to a natural disaster or the threat of one. The control group consisted of participants who did not meet any of the above criteria. Secondly, we created a factor of exposure to threatened safety as a subsample of participants from the disaster group who met at least criterion c), i.e., those who experienced their safety being threatened by a natural disaster. Based on this factor we formed the disaster threat group. Thirdly, we created a factor which specified participants from the disaster group who were exposed to the risk to experience seasonal disasters. The resulting seasonal disaster group was based on participants meeting at least one of the following criteria: a. The participant lives in a town where seasonal natural disasters (such as mudslides and avalanches) have occurred in the past. b. The participant lives in a town where evacuation plans for seasonal natural disasters are available. c. The participant reported that their safety was threatened by seasonal disasters. d. The participant had to evacuate their home due to seasonal disasters or the threat of seasonal disasters. Finally, we formed a factor that selected participants from the seasonal disaster group who met at least criterion g), i.e., those who experienced safety threats due to seasonal disasters. This formed the seasonal disaster threat group.