Survey data on household spatial quality and experiences of stress

This data article describes a dataset of 1,668 cases representing self-reported assessments of housing inadequacy and perceived housing stress. The dataset also contains person-level and household-level demographic data to contextualize the above measures. A second supplemental file contains the text of the survey instrument. Discussion of theoretical background and measures development as well as a more detailed socioeconomic profile of the sample is available in the associated research article http://dx.doi.org/10.1016/j.jenvp.2016.01.002(Campagna, 2016) [1].


a b s t r a c t
This data article describes a dataset of 1,668 cases representing selfreported assessments of housing inadequacy and perceived housing stress. The dataset also contains person-level and household-level demographic data to contextualize the above measures. A second supplemental file contains the text of the survey instrument. Discussion of theoretical background and measures development as well as a more detailed socioeconomic profile of the sample is available in the associated research article http://dx.doi.org/10.1016/ j.jenvp.2016.01.002 (Campagna, 2016) [1].

Value of the data
The raw data of self-reported spatial attributes and psychosocial perceptions may be examined using such statistical methods as analysis of variance, linear regression, factor analysis, or structural equation modeling.
The dataset includes person-level and household-level demographic attributes. The categorical coding of these fields may allow for comparisons of between-group differences from this sample to parallel samples in other similar studies elsewhere.
The self-reported assessments of perceived dwelling quality are defined within a transactional model of person-environment fit. Analyses resulting from data generated by such a conceptual lens may be compared to findings from other datasets collected using alternative theoretical models of housing satisfaction.

Data
Each of the 1668 cases contains the self-reported responses of a single study participant. The data are grouped into fields that recorded individual-level attributes, dwelling-level social and spatial features, and respondents' self-reported perceptions of stress and efficacy. Table 1 summarizes the variables contained in the provided data file.

Experimental design
The dataset contained in the Excel file includes demographic information at the level of the individual and of the household and also self-reported responses to two scales measuring, respectively, household interior inadequacy and perceived housing stress.
Individual demographic data include respondent age, gender, hours employed per week, and several items pertaining to geographic origin. Household-level scale-type data include total household size, number of minor children up to age 17, number of habitable rooms, and number of rooms used for sleeping. Items regarding household composition asked about the presence of seven sorts of immediate family, extended kin, and persons not related to the respondent. Household composition data asked whether any children in the household fell into three specified age ranges. Respondents were also asked about their contribution to household expenses, with four options ranging from paying all accounts to not contributing at all to household expenses. As well, respondents were asked whether they provided a significant amount of caregiving for minor children or for elderly or disabled persons living in the same household.
Scale items measuring housing inadequacy focused on seven spatial attributes of the dwelling interior. Some items asked about physical features such as storage and functionality of utilities and mechanical systems, whereas other items focused on spatial features related to interpersonal interaction. Theses latter items were written to elicit information about dwelling characteristics noted as significant in the environmental psychology literature [2]. Among these desirable features are privacy [3] and spaces for retreat and restoration [4,5]. Respondents first answered to what extent each of seven domestic interior features was available to them and then answered how important each domestic interior feature was to them. The perceived housing stress items were adapted from the ten-item Perceived Stress Scale developed by Cohen and Williamson [6]. The original Perceived Stress Scale is a global measure whereas the rephrased items used here focused the responses on the experience of efficacy and helplessness within the household. Two additional items were added to capture the adverse effects of social withdrawal [7,8] and the mitigating effects of predictability [9].

Materials
Survey data was collected via an anonymous self-administered paper questionnaire. The Word file included in the supplementary material contains the full phrasing of the survey items.

Method
Participants self-selected into this convenience sample. Respondents were community residents enrolled at a junior college in an outer borough of New York City, and all were students in an introductory psychology course at the time of survey administration. As part of a human-subjects approved protocol, respondents could choose among various options for fulfilling a course research participation requirement. During 2012 and 2013, a total of 1885 respondents completed the survey. The present dataset of 1668 cases did not retain 104 where the respondents reported living alone and another 113 where the respondents reported living at their home for less than three months.
For the household inadequacy assessment, the dataset contains three variables for each of the seven attributes described above under Experimental design. The first and second paired sets of items each used Likert-type responses [ Table 1]. The first set of items asked about the availability of each feature, ranging in four intervals from never to always, whereas the second set of items asked about the importance of each dwelling feature to the respondent, ranging in four intervals from not at all important to very important. The availability features were coded to range from À 3 (never) to 0 (always), and the importance features were coded to range from 0 (not at all) to 3 (very). The third set of variables represented difference scores for each corresponding pair of availability and importance items. This third set of variables was computed by subtracting the availability score from the importance score for each attribute's respective pair, hence diff1 ¼want1 -have1. (In the provided data file, the seven computed difference scores are highlighted for clarity.) A difference score of up to three intervals indicated adequacy of the particular feature, but a difference score of greater than three intervals indicated inadequacy of the particular feature. Scale items measuring perceived housing stress used Likert-type responses to elicit respondents' perceptions in five possible intervals ranging from never to very often. Seven of these items measured helplessness and have negative valence, whereas five items measured efficacy and have positive valence. Of the positive-valence items, the provided dataset uses the suffix "o" after the variable label to indicate the original coding and the suffix "R" to indicate the reverse-coded data. The provided dataset highlights the reverse-coded positive-valence items. (Refer to supplementary materials.)