{"id":{"repo_id":"ghent","oai_identifier":"oai:archive.ugent.be:4100284"},"canonical_url":"https://search.dev.ndltd.org/etd/ghent/oai:archive.ugent.be:4100284","repository":{"repo_id":"ghent","name":"Ghent University","base_url":"https://biblio.ugent.be/oai"},"display":{"title":"Response styles in consumer research","abstract":"In researchers’ efforts to better understand consumers, questionnaires are an indispensable source of data. In consumer surveys the Likert item format, where respondents rate their agreement with specific statements, is very popular. However, responses to such items may be biased by response styles, defined as respondents’ tendencies to disproportionately select specific response options. A well-known example is the acquiescence response style, i.e. the tendency to disproportionately use the response options expressing agreement, but respondents may also make disproportionate use of the extreme options, the midpoint option, or the options expressing disagreement. Despite repeated warnings regarding the biasing effect of response styles, most survey research does not control or correct for their impact. A reason for this may be the incomplete understanding of response styles and their antecedents, as well as the difficulties encountered in measuring response styles. The research programme reported in this dissertation aimed to contribute to the understanding of response styles in consumer research by further crystalizing the conceptualization of response styles, by optimizing measurement of response styles, and by explaining the processes that underly response styles. To this end, five empirical studies were carried out. A first study investigated respondents’ understanding of reversed items in questionnaires. Reversed items relate to the same construct as their non-reversed counterparts, but in the opposite direction (e.g. ‘I love to buy new products’ is a reversal of ‘I dislike the purchase of innovations’). This study indicated that responses to items are influenced by the presence of other items that measure the same construct. The exact functional form of this influence is different for reversals and non-reversals, indicating a difference in the way respondents process both types of items. Since this study questioned the validity of reversals for measuring response styles, in the subsequent studies a measurement method for response styles was proposed that captures response tendencies across random samples of items. In a second study, it was shown that response styles are tendencies which are largely stable over the course of a single questionnaire administration. Study 3 established response styles as largely stable tendencies across different questionnaire administrations with a one year time gap in between and using different sets of questions. A fourth study compared response styles across different modes of data collection (self-administered paper and pencil questionnaires, telephone interviews and self-administered online questionnaires). This study showed that there may be differences in response styles across modes of data collection that cannot be detected by the traditional measurement invariance tests. A fifth and final study found two major segments of respondents that differ in the way they satisfice, i.e. economize on the time and effort invested in responding to questionnaire items. One group tends to disproportionately use the midpoint when satisficing. A second group, when satisficing, disproportionately uses the midpoint as well as the negative and positive extremes of the response scale. In sum, though many questions remain unresolved, this dissertation contributes to a better understanding of response styles. More specifically, theory is enhanced by (1) a further delineation of the concept of response styles, which is translated in a proposed operationalization of response styles, (2) evidence in support of the stabililty of response styles, (3) the establishment of response styles as a potential biasing factor in cross-mode comparisons, and (4) a model that captures the relation of response styles to satisficing.","abstract_html":"In researchers’ efforts to better understand consumers, questionnaires are an indispensable source of data. In consumer surveys the Likert item format, where respondents rate their agreement with specific statements, is very popular. However, responses to such items may be biased by response styles, defined as respondents’ tendencies to disproportionately select specific response options. A well-known example is the acquiescence response style, i.e. the tendency to disproportionately use the response options expressing agreement, but respondents may also make disproportionate use of the extreme options, the midpoint option, or the options expressing disagreement. Despite repeated warnings regarding the biasing effect of response styles, most survey research does not control or correct for their impact. A reason for this may be the incomplete understanding of response styles and their antecedents, as well as the difficulties encountered in measuring response styles. The research programme reported in this dissertation aimed to contribute to the understanding of response styles in consumer research by further crystalizing the conceptualization of response styles, by optimizing measurement of response styles, and by explaining the processes that underly response styles. To this end, five empirical studies were carried out. A first study investigated respondents’ understanding of reversed items in questionnaires. Reversed items relate to the same construct as their non-reversed counterparts, but in the opposite direction (e.g. ‘I love to buy new products’ is a reversal of ‘I dislike the purchase of innovations’). This study indicated that responses to items are influenced by the presence of other items that measure the same construct. The exact functional form of this influence is different for reversals and non-reversals, indicating a difference in the way respondents process both types of items. Since this study questioned the validity of reversals for measuring response styles, in the subsequent studies a measurement method for response styles was proposed that captures response tendencies across random samples of items. In a second study, it was shown that response styles are tendencies which are largely stable over the course of a single questionnaire administration. Study 3 established response styles as largely stable tendencies across different questionnaire administrations with a one year time gap in between and using different sets of questions. A fourth study compared response styles across different modes of data collection (self-administered paper and pencil questionnaires, telephone interviews and self-administered online questionnaires). This study showed that there may be differences in response styles across modes of data collection that cannot be detected by the traditional measurement invariance tests. A fifth and final study found two major segments of respondents that differ in the way they satisfice, i.e. economize on the time and effort invested in responding to questionnaire items. One group tends to disproportionately use the midpoint when satisficing. A second group, when satisficing, disproportionately uses the midpoint as well as the negative and positive extremes of the response scale. In sum, though many questions remain unresolved, this dissertation contributes to a better understanding of response styles. More specifically, theory is enhanced by (1) a further delineation of the concept of response styles, which is translated in a proposed operationalization of response styles, (2) evidence in support of the stabililty of response styles, (3) the establishment of response styles as a potential biasing factor in cross-mode comparisons, and (4) a model that captures the relation of response styles to satisficing.","abstract_has_math":false,"creators":["Weijters, Bert"],"institution":"Ghent University. Faculty of Economics and Business Administration","degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":["Geuens, Maggie","Schillewaert, Niels"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2006,"date_issued":"2006","date_published":"2006","updated_at":"2026-07-24T02:22:57Z","subjects":["Social Sciences","response styles","Survey methods","measurement bias"],"languages":["eng"],"rights":["info:eu-repo/semantics/openAccess"],"rights_urls":[],"identifier_entries":[{"key":"dc:identifier","label":"Identifier","values":["https://biblio.ugent.be/publication/4100284","https://biblio.ugent.be/publication/4100284/file/4100290"],"render_values":[{"text":"https://biblio.ugent.be/publication/4100284","href":"https://biblio.ugent.be/publication/4100284","code":true},{"text":"https://biblio.ugent.be/publication/4100284/file/4100290","href":"https://biblio.ugent.be/publication/4100284/file/4100290","code":true}]}]},"links":{"outbound_url":"http://hdl.handle.net/1854/LU-4100284","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Geuens, Maggie","Schillewaert, Niels"]},{"key":"dc:creator","label":"Author","values":["Weijters, Bert"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2006"]},{"key":"dc:publisher","label":"Institution","values":["Ghent University. Faculty of Economics and Business Administration"]},{"key":"dc:type","label":"Dc Type","values":["dissertation","info:eu-repo/semantics/doctoralThesis","info:eu-repo/semantics/publishedVersion"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Social Sciences","response styles","Survey methods","measurement bias"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["eng"]},{"key":"dc:rights","label":"Dc Rights","values":["info:eu-repo/semantics/openAccess"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://biblio.ugent.be/publication/4100284","http://hdl.handle.net/1854/LU-4100284","https://biblio.ugent.be/publication/4100284/file/4100290"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["In researchers’ efforts to better understand consumers, questionnaires are an indispensable source of data. In consumer surveys the Likert item format, where respondents rate their agreement with specific statements, is very popular. However, responses to such items may be biased by response styles, defined as respondents’ tendencies to disproportionately select specific response options. A well-known example is the acquiescence response style, i.e. the tendency to disproportionately use the response options expressing agreement, but respondents may also make disproportionate use of the extreme options, the midpoint option, or the options expressing disagreement. Despite repeated warnings regarding the biasing effect of response styles, most survey research does not control or correct for their impact. A reason for this may be the incomplete understanding of response styles and their antecedents, as well as the difficulties encountered in measuring response styles. The research programme reported in this dissertation aimed to contribute to the understanding of response styles in consumer research by further crystalizing the conceptualization of response styles, by optimizing measurement of response styles, and by explaining the processes that underly response styles. To this end, five empirical studies were carried out. A first study investigated respondents’ understanding of reversed items in questionnaires. Reversed items relate to the same construct as their non-reversed counterparts, but in the opposite direction (e.g. ‘I love to buy new products’ is a reversal of ‘I dislike the purchase of innovations’). This study indicated that responses to items are influenced by the presence of other items that measure the same construct. The exact functional form of this influence is different for reversals and non-reversals, indicating a difference in the way respondents process both types of items. Since this study questioned the validity of reversals for measuring response styles, in the subsequent studies a measurement method for response styles was proposed that captures response tendencies across random samples of items. In a second study, it was shown that response styles are tendencies which are largely stable over the course of a single questionnaire administration. Study 3 established response styles as largely stable tendencies across different questionnaire administrations with a one year time gap in between and using different sets of questions. A fourth study compared response styles across different modes of data collection (self-administered paper and pencil questionnaires, telephone interviews and self-administered online questionnaires). This study showed that there may be differences in response styles across modes of data collection that cannot be detected by the traditional measurement invariance tests. A fifth and final study found two major segments of respondents that differ in the way they satisfice, i.e. economize on the time and effort invested in responding to questionnaire items. One group tends to disproportionately use the midpoint when satisficing. A second group, when satisficing, disproportionately uses the midpoint as well as the negative and positive extremes of the response scale. In sum, though many questions remain unresolved, this dissertation contributes to a better understanding of response styles. More specifically, theory is enhanced by (1) a further delineation of the concept of response styles, which is translated in a proposed operationalization of response styles, (2) evidence in support of the stabililty of response styles, (3) the establishment of response styles as a potential biasing factor in cross-mode comparisons, and (4) a model that captures the relation of response styles to satisficing."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Response styles in consumer research"]}]}],"canonical_facts":{"dc:contributor":["Geuens, Maggie","Schillewaert, Niels"],"dc:creator":["Weijters, Bert"],"dc:date":["2006"],"dc:description":["In researchers’ efforts to better understand consumers, questionnaires are an indispensable source of data. In consumer surveys the Likert item format, where respondents rate their agreement with specific statements, is very popular. However, responses to such items may be biased by response styles, defined as respondents’ tendencies to disproportionately select specific response options. A well-known example is the acquiescence response style, i.e. the tendency to disproportionately use the response options expressing agreement, but respondents may also make disproportionate use of the extreme options, the midpoint option, or the options expressing disagreement. Despite repeated warnings regarding the biasing effect of response styles, most survey research does not control or correct for their impact. A reason for this may be the incomplete understanding of response styles and their antecedents, as well as the difficulties encountered in measuring response styles. The research programme reported in this dissertation aimed to contribute to the understanding of response styles in consumer research by further crystalizing the conceptualization of response styles, by optimizing measurement of response styles, and by explaining the processes that underly response styles. To this end, five empirical studies were carried out. A first study investigated respondents’ understanding of reversed items in questionnaires. Reversed items relate to the same construct as their non-reversed counterparts, but in the opposite direction (e.g. ‘I love to buy new products’ is a reversal of ‘I dislike the purchase of innovations’). This study indicated that responses to items are influenced by the presence of other items that measure the same construct. The exact functional form of this influence is different for reversals and non-reversals, indicating a difference in the way respondents process both types of items. Since this study questioned the validity of reversals for measuring response styles, in the subsequent studies a measurement method for response styles was proposed that captures response tendencies across random samples of items. In a second study, it was shown that response styles are tendencies which are largely stable over the course of a single questionnaire administration. Study 3 established response styles as largely stable tendencies across different questionnaire administrations with a one year time gap in between and using different sets of questions. A fourth study compared response styles across different modes of data collection (self-administered paper and pencil questionnaires, telephone interviews and self-administered online questionnaires). This study showed that there may be differences in response styles across modes of data collection that cannot be detected by the traditional measurement invariance tests. A fifth and final study found two major segments of respondents that differ in the way they satisfice, i.e. economize on the time and effort invested in responding to questionnaire items. One group tends to disproportionately use the midpoint when satisficing. A second group, when satisficing, disproportionately uses the midpoint as well as the negative and positive extremes of the response scale. In sum, though many questions remain unresolved, this dissertation contributes to a better understanding of response styles. More specifically, theory is enhanced by (1) a further delineation of the concept of response styles, which is translated in a proposed operationalization of response styles, (2) evidence in support of the stabililty of response styles, (3) the establishment of response styles as a potential biasing factor in cross-mode comparisons, and (4) a model that captures the relation of response styles to satisficing."],"dc:format":["application/pdf"],"dc:identifier":["https://biblio.ugent.be/publication/4100284","http://hdl.handle.net/1854/LU-4100284","https://biblio.ugent.be/publication/4100284/file/4100290"],"dc:language":["eng"],"dc:publisher":["Ghent University. Faculty of Economics and Business Administration"],"dc:rights":["info:eu-repo/semantics/openAccess"],"dc:subject":["Social Sciences","response styles","Survey methods","measurement bias"],"dc:title":["Response styles in consumer research"],"dc:type":["dissertation","info:eu-repo/semantics/doctoralThesis","info:eu-repo/semantics/publishedVersion"]},"updated_at":"2026-07-24T02:22:57Z"}