Some Psychological Measures for Use in Societal Impacts Research

            A growing number of atmospheric and social scientists have developed an interest in societal impacts research and practices.  This growth is evidenced by training and collaborative networks formed through the Weather and Society Integrated Studies (WAS*IS) initiatives through the National Center for Atmospheric Research (NCAR) and a new forum the for dissemination of research encompassing both the social and atmospheric sciences, Weather, Climate and Society, published by the American Meteorological Society (AMS).  With increasing interest in weather and social science research has come a growing need for measurement tools that can quantify or qualitatively characterize human attitudes, beliefs, emotions, and behaviors in the context of weather and climate.

            The author's purpose for this paper presentation is to introduce several new psychological measures for societal impacts research and also to illustrate their performance and use in several research projects.  The measures include: 1. The Weather Salience Questionnaire (WxSQ), which has been published previously (Stewart 2010) and now has a new, 7-item short-form, 2. the Weather Risk-Taking Scale (WRTS), 3. the Fear of Weather Survey (FOWS), and 4. the Hurricane Response Self-Efficacy Scale (HuRSES).  These measures assess attitudes, awareness, emotional and motivational processes associated with routine and severe or extreme weather events.  It is hoped that by exhibiting these measures that other atmospheric and social scientists attending the AMS annual conference will learn about their performance and availability for research projects that they may be designing.

            A description of the nature and psychometric/statistical performance of the measures (excepting the HuRSES) appears in Table 1.  A brief description of the measures follows.  The WxSQ assesses seven different facets of the ways in which weather is important or matters to people.  People exhibiting higher scores on the WxSQ are both more affected by and responsive to the weather.  The WRTS measures three dimensions that relate to the extent to which people take risks with the weather (e. g., driving over a flooded roadway).  These dimensions are: a. perceived costs in taking the risk, b. perceived benefit of taking the risk, and c. likelihood of weather risk-taking.  The WRTS is an important tool in assessing the extent to which people may put themselves in harm's way during extreme or severe weather.  The FOWS measures the kinds of weather that may result in fear and anxiety in people and quantifies the degree of this fear.  The FOWS consists of a total score, with higher scores indicating greater degrees of fear of different weather events.  Study of fear of weather is important as up to 10% of the United States population may have a significant fear of weather; in addition, fear of weather may especially motivate people to take precautions ahead of severe weather.  The HuRSES assesses the extent to which people believe that they can personally prepare for a nearby hurricane and, if necessary, evacuate to a hurricane shelter.  This measure is important because it may help explain, in part, the variability that exists in coastal residents' responsiveness to approaching tropical cyclones.

            For each of these measures the author will: 1. Briefly characterize its development, nature, purposes and uses, 2. the psychometric properties of the measure, and 3. empirical results that have been obtained in using the measure thus far.  This information is provided in summary form in Table 1.  Finally, the author will describe the process for obtaining and using the measures, which are available on his website. 

            Table 1 – Summary Characteristics of Psychological Measures for Use in Societal Impacts Research

Measure	Purpose	Theoretical Basis	Number of Items	Subscales	Sample Size	Measure Reliability*	Relationships with Other Variables
Weather Salience Questionnaire – Short Form (WxSQ-SF)	Measure the extent to which people place importance on weather and its effects.	Derived from work on a related construct, environmental salience (Stokols, 1979)	7	None. A single scale provides a brief indication of weather salience.  Higher scores denote greater importance placed upon the weather	1,465	0.70	a. Correlation with full, 29-item WxSQ, r = 0.84 b. WxSQ-SF differed according the differences in the respondents' climatic zone. c. WxSQ-SF was significantly correlated with peoples' sources and uses of weather information. d. WxSQ-SF was significantly correlated with degree of confidence in forecasts
Weather Risk-Taking Scale (WRTS)	Assess the perceived costs, benefits, and likelihood of taking weather-related risks.	Developed using the rationale for Domain Specific Risk-Taking (DOSPERT; Blais & Weber, 2006)	96	a. Perceived Benefits b. Perceived Costs c. Likelihood of Taking Higher scores indicate greater benefits, costs, and likelihood of risk-taking.	246	0.87 to 0.93	Likelihood of weather risk-taking was negatively related to: a. perception of weather risk b. perception of health and recreational-based risks c. fear of weather (FOWS) Likelihood of weather risk-taking was positively related to: d. perceived benefits of weather risks. e. perceived benefits of health and recreation risks
Fear of Weather Scale (FOWS)	Quantify the nature and degree of weather-related fears that people express.	A general measure of fear in psychotherapy clients, the Fear Survey Schedule (Lang & Lazovik, 1963)	89	None. A single scale provides an overall assessment of the degree of fear of weather.  Higher scores indicate greater fear.	1,500	0.96	Fear of Weather Scale (FOWS) scores were significantly correlated with weather salience, weather-related risk-taking, and fears of other natural hazards (e. g., earthquakes, tsunamis).
Hurricane Response Self-Efficacy Scale (HuRSES)	Measure the extent to which people feel like they can personally prepare for a hurricane.	Derived on the basis of Bandura's (1997) self-efficacy theory.	10	a. Outcome Expectations b. Efficacy Expectations	≈ 200	In process	In process – Study is currently underway to relate the HuRSES to the WxSQ, FOWS, and WRTS.
*Note: Scale reliability was assessed using Cronbach's coefficient alpha (α). Alpha ranges from 0 to 1, with higher scores indicating greater internal consistency (and thus reliability) of the component items.