Effects of dress, ethnicity, and gender on latency to service in retail stores.

Abstract

The purpose of this study was to investigate the degree to which appearance, ethnicity, and gender influence how promptly individuals are offered customer service in retail environments. The participants consisted of salesclerks in 84 one-entrance, service-oriented retail stores located in two large shopping malls in the midwest. The stores were assigned to one of three categories: "male"-stores selling primarily male-oriented goods, such as electronics or sporting goods; "female"-stores selling primarily female-oriented goods, such as clothing or beauty aids; and "gender neutral" stores selling primarily gender neutral goods, such as jewelry or home furnishings. Three men (one African-American, one Anglo-American, one Hispanic-American) and three women (one African-American, one Anglo-American, one Hispanic-American) their 20s served as confederates in the conduct of the experiment. Two styles of dress, business and casual, were adopted by each confederate. The assignment of confederates and attire to type of store and shopping mall was random with the restrictions that: (a) each confederate was to participate at each mall in only one style of clothing (i.e., if the African-American male wore casual clothing in Mall 1. he wore business attire Mall 2), (b) casual and business clothing be worn equal number of times at each mall. Each confederate carried a silent, digital stopwatch in the palm of his or her band. The stopwatch was activated as the confederate crossed the threshold of each store if an available salesclerk was in plain view or as soon as an unoccupied salesclerk became available. The confederates reported the latency to serve data to a data recorder positioned in the mall walkways outside the stores. A preliminary analysis compared the latency to serve data (seconds) between two shopping malls. As the results of this analysis failed to yield a significant difference, !(334) = 1.31, p> .05, the malls were considered equivalent and this factor was not evaluated in subsequent analyses. A four-factor analysis of variance (ANOVA) incorporating gender (male-female), race (African American-Hispanic American-Anglo American), type of dress (businesscasual), and type of store (male-female-gender neutral) was used to analyze the latency serve data. This analysis yielded significance for the race, F(2, 468) =85.44, p < .001, type of dress, F(1, 468) = 33.18, p < .001, gender x dress, F(l,468) = 24.05, p < .001,and gender x race x dress, F(2, 468) =5.95, p = .003, effects. The Newman-Keuls procedure was used to probe the significant sex x race x dress interaction. The results of these contrasts indicated that regardless of type of dress the Anglo-American man and woman were waited on significantly faster (p < .01) than all other conc1itions, except the Hispanic-American man wearing business attire. The latency to serve the Anglo-American man and woman and Hispanic-American man wearing business clothing did not differ significantly. Additional significant comparisons indicated that the African-American woman wearing either business or casual clothing, the African-American man wearing casual clothing and the Hispanic-American man wearing casual clothing received significantly (p < .01) slower service than did the African-American man dressed in business clothing, the African-American woman dressed in either business or casual clothing, and the Hispanic-American man dressed in casual attire. All other contrasts were not statistically significant. The results of this study indicated salesclerks form unjust impressions of their customers based on easily observable characteristics including gender, dress, and most importantly their race. These initial impressions, in tum, drastically affect how promptly minority customers are serviced in retail environments. The findings support the essentially negative evaluation of African-American and Hispanic-American individuals our country, with these customers experiencing longer response times from salespersons than their Anglo-American counterparts. Further research might also be extended to include other ethnic groups such as Caribbean-Americans, European-Americans, or Asian-Americans, which appear to be growing segments of our population.