disorder and conflict, feel-good factor, social capital, social cohesion, sport participation
This study tests the construct validity of two different social impact scales by comparing the perceived social impact of a non-mega sport event for the same group of respondents. To date, several theories have underpinned the development of various social impact scales, and there is a need for more robust and unified measurement tools. Data were collected from 626 residents in the context of the 2014 Ontario Summer Games (OSG). Event attendees (29%) and non-event attendees (71%) completed a questionnaire (electronically or on paper) which included 17 social impact items, reflecting two previously developed social impact scales (SIS-A consisting of 4 constructs and SIS-B consisting of 5 constructs). Principal Component Analysis showed sufficient convergent validity of theoretical constructs in both scales, but Pearson correlations between the constructs only partially supported discriminant validity. Therefore, EFA was conducted revealing two components: a “positive” and “negative” social impact factor of SIS-A (whether or not a new sport participation variable was included). EFA of SIS-B without the sport participation variable resulted in a similar set of two components. However, when the sport participation variable was added to SIS-B, three components appeared. The positive social impact was now represented by two constructs; the third factor remained the negative social impact factor. An EFA of all 17 items revealed the same three constructs: (1) “Feel-Good and Social Cohesion”, (2) “Social Capital”, and (3) “Conflict and Disorder”. Based on the findings, a scale of 13 items and three constructs is proposed.
Taks, Marijke; Meagan, Littlejohn; Wiood, Laura; and Snelgrove, Ryan. (2015). Construct validity of social impact scales for sport events.