Evaluation of the Nextdoor KIND Challenge

The EUROHIS-QOL-8 is an 8-item measure of quality of life consisting of questions that assess overall QOL, general health, energy, daily living activity, self-esteem, social relationships, finances and home. The measure is derived from the WHOQOL-BREF and shares a similar 5-point Likert-scale response format. The scale has demonstrated good qualities in term of internal consistency (Cronbach's alpha = 0.83) and satisfactory convergent and discriminant validity (Schmidt et al., 2005). The range of possible scores is 8-40 where higher scores indicate higher quality of life.

The PHQ-8 is an 8-item measure of depression severity based on 8 of the 9 item criteria of the DSM-IV. The measure was derived from the PHQ-9 however; the ninth question was omitted from this study as depression was not a primary outcome. Depression severity is scored based on the presence of depressive symptomology in the previous 2 weeks of measure admission. The scale uses a 1 (not at all) to 4 (nearly every day) Likert scale with higher scores indicating higher levels of depressive symptomology. The range of possible scores is 8-32. The PHQ-9 has been shown to have good criterion and construct validity and excellent internal consistency (Kroenke, Spitzer, \& Williams, 2001). The PHQ-8 shares similar properties in terms of validity and reliability and is therefore an adequate alternative to its 9-item scale counterpart (Kroenke \& Spitzer, 2002).

The Mini-SPIN is a brief 3-item measure of generalised social anxiety disorder. The measure employs a 5-point Likert scale form 1 (not at all) to 5 (extremely), with higher scores indicating a greater level of generalised social anxiety. The range of possible scores is 3-15. The Mini-SPIN has shown high sensitivity in detecting social anxiety disorder (Connor et al., 2001).

This four-item measure of stress was derived from the original 14-item Perceived Stress Scale (PSS). This measure consists of both positively and negatively worded items that asses an individual's evaluation of stressful events. The measure employs a 5-point Likert scale ranging from 0 (never) to 4 (very often). The range of possible scores is 0-16 where higher scores indicate higher levels of perceived stress. The PSS-4 has been shown to negatively correlate with levels of perceived health, social support, being male, and older age (Warttig, Forshaw, South, \& White, 2013). The scale has demonstrated fair reliability and adequate psychometric properties (Warttig et al., 2013).

The PANAS-SF is a 10-item short form version of the Positive and Negative Affect Scale scale. The measure asks respondents to rate the extent to which they feel a particular emotion along a 5-point Likert scale ranging from 1 (very slightly) to 5 (extremely). The range of possible scores is 5-25 for each subscale (positive affect and negative affect) where higher scores indicate higher levels of positive or negative affect. This subscale measure has been shown to have high levels of internal consistency (Watson et al., 1988) and acceptable levels of convergent validity (Roesch, 1998).

The Social Capital Scale is a 7-item measure designed to assess social cohesion and trust in a community. The scale was derived from a similar scale used in Sampson, Raudenbush, and Earls (1997). This measure consists of both positively and negatively worded items and is recorded on a 2 point Likert scale ranging form 0 (strongly disagree or disagree) to 1 (strongly Agree or agree). The range of possible scores is 0-14 where higher scores indicate higher levels of social cohesion and trust. This measure was shown to be a reliable tool to measure social cohesion and trust at a neighbourhood level (Sampson et al., 1997).

This question asks if participants thought their neighbourhood was improving from 1 (improving) to 3 (declining).

This question asks how important participants thought it was to know their neighbours from 1 (very important) to 5 (very important)

The 3-item scale was modified from the Social Relationship Index (34). Three factors measuring support, ambivalence, and aversive. We ask how participants feel towards their neighbours when needing advice, understanding, or a favour - helpfulness, upsetting, mixed/conflicted feelings on a Likert scale from 1 (not at all) to 6 (extremely).

Participants will be asked to reflect on their interactions with their neighbours in the past month using a dichotomous yes/no to the following: absence/presence of neighbour conflict, critical comments towards the participant, participant critical of neighbours.

This item asks the number of people the participants knows in their neighbourhood, 0 (0-5 neighbours) to 4 (15+ neighbours).

We measured acceptability around these factors, including how connected (1 not at all connected to 10 very connected), how meaningful was the activity (1 not very meaningful to 10 very meaningful), how safe they felt when completing the challenge (1 not very safe to 10 very safe), how positive they felt (1 not very positive to 10 very positive), how comfortable they felt doing the activity (1 not at all comfortable to 10 very comfortable). It is anticipated that the Nextdoor KIND challenge will yield ratings of more than 5 indicating higher levels of acceptability across these outcomes.

Feasibility will be assessed by a high retention rate (i.e., <40% drop-out).

The LSNS is an 18-item scale that assesses the frequency and quality of contact - such as talking about private matters - in an individual's network. There are three subscales and each consists of 6 items relating to family, neighbour and friend connections. The scale employs a 0 (none) to 5 (nine or more) Likert scale and includes 6 items (eg. How many relatives do you see or hear from at least once a month?). Higher scores indicate larger social networks and lower risk of social isolation. The scale has demonstrated adequate levels of reliability and the proposed clinical cut-points showed good convergent validity (Lubben \& Gironda, 2003). The range of possible scores is from 0-90 where higher scores indicate more social connections and lower risk of social isolation.

We assess for unintended harms (e.g., conflict) during this period. In this case, we measure the occurrence of neighbour conflict with a single dichotomous item, 'Did you have a conflict with a neighbour because of the KIND challenge'. This item is measured with a yes or no response.