Notes on Sutton and Rafaeli

The purpose of these notes is to map the Sutton & Rafaeli article onto the categories discussed in class, such as the sections of a research report, and also things like independent and dependent variables, intervening variables, etc. Of course, it should be noted that this article has an unusual structure, in that it is really two studies back-to-back.

The exact reference for the article is:

• Sutton, R.I. and A. Rafaeli. 1988. "Untangling the relationship between displayed emotions and organizational sales: The case of convenience stores." Academy of Management Journal. 31(3):461-487.

Title

The title gives a description of the research question (how expression of emotions by employees affects customers) and also tells what the general research site is (a convenience store).

Introductory Section

This paper wraps up the literature review and tje statement of the research question into a single unlabeled section (the very beginning). The next section, entitled "displayed emotions as organizational attributes", contains the theoretical framework.

Unlabeled Section

• Starts by explaining what has been researched about emotions in organizations to date.
• Introduces the distinction between emotions felt and emotions displayed.
• Orgs have normative rules about what should be displayed
• Can achieve uniformity by hiring selection process, and by training, and rewards/punishments
• Orgs control emotional display because these displays manipulate customers
  • although not a fact, this is the observation that the paper seeks to explain

Displayed Emotions as Organizational Attributes

• customers only interact with boundary spanners, but take away impression of whole organization
  • customers do not interpret employees behavior as just individual behavior: it reflects on the org
• other researchers have seen it this way too
• that warm displays promote sales could be an implication of learning theories
  • smile reinforces customer, who keeps buying in order to get that smile again
  • this can also occur through vicarious learning

Stop and think: what other theories could account for displayed emotions affecting sales?

Methods

This is the longest section. It describes how the study came to be, what the organizational context was, and what how the variables were measured.

Research site

• Management interested in improving customer service as a point of competitive advantage
• Had training, and reward system based on lottery
• Wanted quantitative evidence that this would work

Sample

• Sampling unit is the store
• Unit of observation is the transaction

Procedures

• Observers acted like customers, surreptitiously watching interactions between clerk and customer

Independent Variable

• Looked at two verbal items (greeting and thanking) and two non-verbal (smiling and eye contact)
• Decide in advance what counts as what. these are behavioral codes
• Observers recorded presence/absence of code
• For each store, compute % of transactions that had the codes
• Test whether it makes sense to look at this at the store level (versus individual level)

Control Variables

Control variables are variables that are potentially correlated with both the independent and the dependent variables and which you have no physical control over (because it is field study, not a lab experiment).

• Sex of clerks (% female)
• Sex of customers (% female)
• Professionalism of clerk's dress
• Store's stock level
• Line length
• Owned by corporation or franchiser
• Store supervision costs
• Region

Dependent Variable

• Total store sales for 1984 (expressed as z-scores for confidentiality)

Results

• Displayed emotion was negatively related to store sales
• Contradiction of expectations

Intro to the Second Paper

• The need for more exploratory work to develop new understanding of relationships among variables

Qualitative Methods

• Long observations
• Interviews with managers
• Clerk for a day

Revised Perspective

• Over time, customers and clerks work out a pair of cognitive scripts for what happens at the store, one for busy times and one for slow
• Store pace is the cue for switching norms of expressed emotions

Busy Times

• Customers are inputs to be processed rapidly
  • Both clerks and customers want the transactions to proceed as quickly as possible
  • Being nice does not take time, but can encourage customers to hang around and talk
  • No-nonsense attitude is less irritating to those waiting in line behind current customer (perceived efficiency)
• Busy pace stresses the clerks (and customers), so harder to maintain (fake) good cheer

Slow Times

• No pressure to be quick
• Customers come in with different attitude when its slow
• Clerks glad to see people when it's really slow -- less boring

Re-analysis: Methods and Results

• positive emotion is dependent variable
• store sales and line length (measures of pace) are independent variables
• Individual-level analyses
  • length of line predicts display of positive emotions
  • recurring high pace (store sales) should decrease the effect of line length on displayed emotion, because clerks in normally slow stores get upset more easily: and it does.
  • [but the theory of cued script switch suggests the opposite: the authors don't notice that]

Discussion

Things To Think About

1. Store pace can be seen as a moderating variable which decreases the effect of displayed warmth on sales. Warmth is good for sales when business is slow, but it is bad for sales when business is fast.
   • the authors conclude that sales determines warmth, not the other way around. Does it have to be just one or the other?
2. What really is the unit of analysis? Who displays emotion -- organizations or people? What would be a more careful model that does not confound levels of analysis?