Replication data and code for: Decomposing Affective Response to Personalized Email in Online Travel Services

Item-level survey responses, variable dictionary and analysis code underlying the article "Decomposing Affective Response to Personalized Email in Online Travel Services". The response file contains 250 complete responses to a cross-sectional questionnaire administered to customers of a Tunisian online travel agency, drawn from its family-with-children segment: a sequential respondent identifier, 42 five-point Likert items covering twelve reflective constructs, four demographic variables, and thirteen legacy construct-score columns that are not used in any reported result. No names, email addresses or other direct identifiers are present. A variable dictionary documents every column: construct, item number, administered wording in English translation, response format and coding, observed range and descriptives. A standalone Python script, requiring only numpy and scipy, takes the response file as its only argument and regenerates every table in the article and in its supplementary material. The README states in full the conventions a replication must match, together with the software versions and random seeds of the reference run. Participation was voluntary and anonymous; the questionnaire opened with a statement of the study's purpose, its academic character and the anonymity of responses. No special-category personal data were collected.