A convergent validity test within the payment card format using simulation techniques

• Published in: Applied economics Vol. 51; no. 34; pp. 3770 - 3786
• Main Authors: Voltaire, Louinord, Jeanty, P. Wilner, Pirrone, Claudio, Mahieu, Pierre-Alexandre
• Format: Journal Article
• Language: English
• Published: London Taylor & Francis Ltd 21.07.2019; Taylor & Francis (Routledge)
• Subjects: Bias; Convergent validity; Credit cards; Economic analysis; Economic theory; Humanities and Social Sciences; Respondents; Simulation; Validity; Willingness to pay; payment card versions; convergent validity; simulation techniques; Contingent valuation
• ISSN: 0003-6846; 1466-4283
• DOI: 10.1080/00036846.2019.1584382

A convergent validity test is performed between two groups of versions of the payment card format. The first group, the classic payment card (CPC), asks respondents to report their willingness to pay (WTP) as a point from a list of amounts, and then treats each WTP response as an interval. The second group generates WTP data that may contain both single point and interval values. It includes the two-way-payment ladder (TWPL) (respondents have to tick amounts they would definitely pay and cross amounts they would definitely not pay), and point-interval payment card (PIPC) (respondents have to tick their WTP as either a point or an interval). The test is conducted using data from one TWPL study and two PIPC studies. For each study, we use WTP values stated to simulate 200 CPC WTP datasets, which allows controlling for any behavioral biases likely to confound the outcome of the test. The results challenge the conventional way of eliciting WTP under the CPC and the typical way of treating WTP data from this PC version. Although convergent validity holds between the TWPL/PIPC and CPC, parameter and mean WTP estimates from the TWPL and PIPC are more stable as PC intervals widths increase.