Friday, July 25, 2008

An experiment in the Czech Republic

Anecdotal evidence: A bilingual friend of other travellers reported that she felt she received better service when speaking English than German in Prague.

Hypothesis: Service workers in Prague give better treatment to tourists who speak English than those who speak German.

Introduction: It was previously reported (unpublished data) that service varies in The Czech Republic based on the language used. While it is tempting to believe that this is the case, a lack of formal, controlled experimentation means that no definite conclusions can be made.

Prague is a worldwide tourist destination. Virtually untouched by war, the city is a beautifully preserved example of architecture of all periods from Medieval times to modern day. The city is dense with tourists and a gigantic tourism industry has grown out of this. In Wenceslas Square, in particular, many fast food stands make particularly good business serving various American and German-style fast foods all along its length.

This report details the investigation of the effect of language on service in the Czech capital city, Prague. Wenceslas Square was chosen as the experimental site due to the density and ease of availability of the same product in multiple locations in close proximity.

Materials and Methods: One hungry traveller looked for something to eat, and stopped at two stands consecutively for the cheapest fast food offered, the Hamburger (25 Kc). At one stand the experimenter requested "one hamburger, please" and at the other "einen Hamburger, bitte" and recorded the results in terms of service and food quality.

Sample size: 1

English: researcher received polite treatment, and an outstanding hamburger, complete with ketchup, lettuce and a perfectly toasted bun.

German: The experimenter received rude, rushed service, and hamburger was hastily thrown together, with a soggy bun and overall poor quality.

Relative quality of service ratio (E/G)*: 100

95% Confidence interval: Relative risk of approximately 0.1 - 10000

*Subjective judgement of the experimenter

Discussion: It was found that service was much better when English was the language employed, but unfortunately with the sample size employed, the difference could not be considered significant. Further studies are required to support the findings of this report. Nevertheless the findings are encouraging, and are what would be expected if the hypothesis is, indeed, true.

It should be noted that in both cases, the servers replied in English, regardless of the language the meal was ordered in, indicating a bias in service, and this could not be controlled for.

In addition, it was discovered afterward that although payment was made in the form of a 50Kc bill at both stands for a 25Kc meal, at the end the experimenter had only 25 + 15Kc in change, meaning that one stand did not return the appropriate change. This requires further investigation, as it is not known if this discrepancy occurred when the English or German transaction occurred.

Unfortunately, although both stands were owned and operated by the same company, location was a confounding factor as the "German" stand was noticeably busier than the other.

A limiting factor in this experiment was that the experimenter does not speak any Czech, and so an important control could not be conducted. However, after two Czech-style hamburgers were consumed, there was little desire for another and so this control would likely have gone unperformed anyway.

Conclusions: Although no significant results were obtained due to the limited sample size, the results support the hypothesis that English-speaking tourists do in fact receive better treatment in Prague than German-speakers.

Recommendations: Further testing is required. The sample size employed did not allow any definite conclusions to be drawn, although the observations recorded are encouraging with respect to the hypothesis. A randomized trial with multiple subjects and locations would prove very informative in establishing the presence or absence of an effect.

No comments: