Multicultural Service Settings-Elmadag et al 2026
Description
Data Set Description This study uses a cross-sectional survey dataset collected from consumers in two countries, the United States and Turkey. The data were gathered through an online consumer panel (Prodege) using a Critical Incident Technique (CIT) approach, where respondents recalled and described a real multicultural service encounter involving immigrant customers. After applying strict data quality screening, including attention checks, completeness, and quality of incident descriptions, the final dataset includes 672 usable responses (375 U.S., 297 Turkey). Key Variables The dataset includes multi-item, validated scale measures for: Xenophobia (predictor) Prejudice toward immigrants’ country of origin (mediator) Emotional responses during the encounter Positive emotions Negative emotions Service outcomes Shopping stress Word-of-mouth intentions Cultural context (country: U.S. vs Turkey as proxy for individualism vs collectivism) All measures were adapted from established scales and assessed using Likert-type items. Data Collection Design Method: Survey with embedded critical incident recall Design: Cross-sectional Context: Multicultural service encounters (customer-to-customer interactions) Procedure: Xenophobia measured first Respondents recalled a specific intercultural service encounter Post-incident measures captured emotions, prejudice, and outcomes Data Quality and Processing Removal of: Failed attention checks Incomplete responses Translation handled via back-translation procedure for cross-language equivalence
Files
Steps to reproduce
Data Collection and Generation Process The dataset was generated using a cross-sectional survey design combined with the Critical Incident Technique (CIT) to capture real-world multicultural service encounters. Sampling and Recruitment Participants were recruited through a managed online consumer panel (Prodege) in two national contexts: the United States and Turkey. These countries were selected to represent contrasting cultural orientations, specifically individualistic and collectivistic contexts. Final usable sample: 672 respondents United States: 375 Turkey: 297 Data Collection Procedure Data collection followed a multi-stage structured workflow to reduce bias and ensure temporal separation: Pre-measurement of xenophobia Participants first completed items measuring xenophobia to avoid contamination from recalling emotional experiences. Critical Incident Task (CIT) Respondents were instructed to recall and describe a specific service encounter involving immigrant customers, either through direct interaction or close proximity. They provided: Service context Description of the interaction Explanation of why the incident was memorable Post-incident evaluation After recalling the event, participants completed measures assessing: Prejudice toward the immigrant customer’s country of origin Emotional responses (positive and negative) Behavioral outcomes (word-of-mouth, shopping stress) Measurement Instruments All constructs were measured using established multi-item scales adapted to the service encounter context: Xenophobia: Van der Veer et al. (2013) Prejudice: Stephan et al. (1999) Positive and negative emotions: Watson et al. (1988) Shopping stress: Baker and Wakefield (2012) Word-of-mouth: Zeithaml et al. (1996) All items were measured using Likert-type scales. Data Quality Protocols To ensure reproducibility and reliability, the following screening procedures were applied: Removal of respondents who: Failed attention checks Submitted incomplete surveys Provided low-quality or irrelevant incident descriptions Critical incident quality criteria required: Minimum word count Clear relevance to a service setting Sufficient detail for interpretation Cross-Cultural Instrument Handling A back-translation protocol was used to ensure measurement equivalence: Initial translation into Turkish by two bilingual translators Reconciliation into a single version Back-translation into English by a separate translator