From the literature review (Chapter 2), it is evident that there is a lack of theoretical guidance around store atmospheric stimuli and how retailers might use these to encourage healthier food choices. To better understand this relationship, an important initial step is to observe what is currently employed within the marketplace. This is key, as store atmospheric stimuli have communicative power. They can construct and communicate meaning for the shopper, as highlighted in the literature review (Chapter 2). If store atmospheric cues have the ability to communicate different meaning, store atmospherics in a health-orientated store should, therefore, communicate meanings
related to health. As Ward et al. (1992) highlights, “consumers rely on external and internal environmental cues to categorize (and thus make inferences about) the products and services offered by retailers” (p. 195). Thus, it is assumed that more-healthy food stores would have different store atmospheric stimuli (e.g. internal and external environmental cues) from less-healthy food stores. The observation of more-healthy and less-healthy food store environments in Study 1 thus allows for the development of a comprehensive set of atmospheric stimuli that represent both more- and less-healthy food stores within New Zealand.
Physically observing store atmospheric stimuli in retail stores is not a new method of investigation; Parsons (2011) successfully implemented this method to gain insights into sensory stimuli components that were being used to create store atmospheres. Though Parsons (2011) did not use this method to distinguish between more- versus less-healthy food store environments, the method provides insights into different sensory stimuli components appropriate for this investigation.
Study 1 employs a semi-structured observational approach utilising both subjective human senses and objective measurement devices in natural settings. Before selecting a semi-structured approach, unstructured and structured approaches were taken into consideration. Boote and Mathews (1999) suggested that unstructured observation is typically used in the earlier stages of research (prior to hypothesis development). The current study research question emerges at the early stages of work in the retail literature, but not so early that there is no guidance in the literature. For instance, a pre- existing classification of store atmospherics (e.g., olfactory, auditory, tactile and visual, Kotler, 1973) was identified in the previous chapter enabling guidance to satisfy the research objective. Completely unstructured observations are thus unwarranted with this pre-existing classification. If purely structured observation were utilised, a strictly defined classification system would have been implemented to dictate what to observe, record, and analyse (Gilbert & Churchill, 1999). A drawback of structured observation, however, is that new and additional information cannot be added. Important information might have been missed. A semi-structured approach thereby incorporates the best of both structured and unstructured methodologies. Biases are reduced and reliability increased via the structured approach (Gilbert & Churchill, 1999) and a higher level of validity achieved from the unstructured approach, with new and relevant information added where necessary (Gilbert & Churchill, 1999).
After selecting a semi-structured approach, disguised observation was considered, but rejected. Disguising oneself is generally employed to ensure the observed behaviour is naturalistic (Dodd, Clarke, & Kirkup, 1998). The focal point of this observational study was not shopper behaviour or staff action, however, but the presence and the level of store atmospheric stimuli (Parsons, 2011). Observations in a natural setting can enable realistic data to be collected, as opposed to a contrived or isolated setting. A contrived setting such as the laboratory does not meet Study 1’s objectives to investigate store atmospherics employed in physical stores within the New Zealand marketplace.
To ensure carrying out observations of the retail environments were the correct method to apply, the following six criteria adapted from Boote and Mathews (1999) and McDaniel and Gates (2010) needs to be met:
(1) The phenomenon under investigation needs to be easily observable.
(2) The phenomenon under investigation needs to occur at a subconscious level. (3) The consumers under investigation needs to be either unable or unwilling to communicate directly with the researcher.
(4) The phenomenon under investigation needs to be a social process or a mass activity.
(5) The phenomenon under investigation needs to be repetitive, frequent, or in some manner predictable.
(6) The phenomenon under investigation needs to be of a relatively short duration.
Store atmospheric stimuli (the phenomenon) were easily observable (criteria 1). Even though store atmospherics are known for their non-conscious influence on shoppers (e.g. Knoferle et al., 2012; North et al., 2016; Yalch & Spangenberg, 2000) (criteria 2), they are not “non-conscious” themselves and can still be directly observed and measured (Parsons, 2011) (criteria 1). Store atmospheric stimuli are designed to be sensed (Parsons, 2011) and can be consciously processed, if required (criteria 1). Store atmospherics are unable to be verbally communicated (criteria 3) but can be measured. Retail environments do not operate without store atmospherics and therefore a pattern
of activity can be investigated across the environments, and at scale (criteria 4). Instead of waiting for an event to happen, stores can be visited anytime during opening hours and the stimuli observed; this predictability and frequency allows the researcher ample time to observe and record the atmospheric stimuli (criteria 5). Due to the relatively small size of retail food environments in New Zealand, it would take no longer than an hour to navigate a store when no purchasing decisions are involved. This amount of time is fairly short, making observational data collection practical (criteria 6). Having met the criterion for observations, it seems practical and applicable that this method should be used to help answer the research question of which store atmospherics are representative of a healthier store environment.