2015年-FCA英国金融行为监管局_ms14_6_2_ccms_annex_3_2_39页_1mb

Credit Card Market Study: Consumer Survey Summary

Core Content Overview

This technical report outlines the methodology and design of the FCA's Credit Card Market Study (CCMS) consumer survey, conducted from 1 to 16 April 2015 by YouGov. The survey aimed to build an evidence-based understanding of the UK credit card market, including consumer behavior, attitudes towards different types of credit cards, and the impact of credit card usage. The report includes details on questionnaire development, sampling techniques, data collection, and weighting procedures.

Main Survey Objectives

• Estimate the size of the total credit card market in the UK, focusing on active card holdership
• Enhance the FCA's understanding of consumer behavior related to credit cards
• Investigate usage of specific types of credit cards:
  • Rewards credit cards
  • Cards for no or poor credit history
  • Cards used for balance transfers

The survey targeted approximately 40,000 interviews with UK adults aged 18 and over, with no more than 40% of non-users in the final sample.

Survey Design and Structure

Questionnaire Development

• The questionnaire was developed in collaboration with the FCA and an independent consultant, John Leston.
• It was tested through:
  • An omnibus survey (20–22 February 2015) to estimate population eligibility for different modules.
  • A pilot test (16–25 March 2015) involving online and telephone depth interviews to refine the questionnaire and module allocation rules.

Modular Approach

• The survey was structured into modules to avoid excessive length (maximum 16 minutes).
• Modules were designed to focus on specific aspects of credit card usage and behavior:
  • Credit card held/usage modules:
    • Rewards credit card users
    • No/poor credit history card users
    • Balance transfer users
  • Considering and taking out credit card modules:
    • Took a credit card without considering others
    • Considered and took a credit card
    • Considered but did not take a credit card

Module Allocation

• Respondents were allocated to modules based on their eligibility, determined by their main credit card usage.
• Allocation rules were adjusted after the first stage of the survey to ensure accurate targeting of modules.
• A weighted allocation was applied to ensure the sample was nationally representative.

Sampling and Data Collection

Sampling Strategy

• The survey aimed to achieve a nationally representative sample based on:
  • Age and gender
  • Region
  • Social grade
• Quotas were defined using Mid-2013 ONS data.
• A two-panel approach was used:
  • YouGov panel: 25,506 valid interviews
  • Research Now panel: 14,331 valid interviews

Sampling Process

• Turbo sampling was used by YouGov to dynamically allocate respondents to the most appropriate survey based on demographic needs, time remaining, and other factors.
• Research Now used a similar system but adjusted daily instead of every 30 minutes.

Data Cleaning and Weighting

• Data cleaning was performed to remove:
  • Duplicate responses
  • Erroneous or fraudulent data
• Data editing involved removing some answers for limited numbers of respondents to improve data accuracy.
• The final sample was weighted using:
  • Standard weighting to ensure national representativeness
  • RIM weighting to align with Mid-2013 ONS data
• Weighting factors were calculated and applied to each respondent, as shown in Table 18 and 19.

Strengths and Limitations

• Strengths:
  • Nationally representative sampling
  • Modular design to reduce survey length
  • Rigorous pilot testing and data cleaning
  • Independent statistical oversight by Derek Farr
• Limitations:
  • Potential for data discrepancies
  • Challenges in achieving exact quota targets
  • Limited ability to capture all credit card-related behaviors due to module exclusivity

Key Tables and Figures

• Table 1: Research timeline with key project stages
• Table 2: Allocation of respondents to questionnaires
• Table 3: Eligibility criteria for credit card held/usage modules
• Table 4: Eligibility criteria for considering and taking out credit card modules
• Table 5: Omnibus survey estimates of population eligibility
• Table 6: Main survey quota targets
• Table 7: Final achieved interviews vs. quota targets
• Table 8: Differences in profiles between single and multiple card holders
• Table 9–12: Module allocation for single and multiple card holders
• Table 13–15: Eligibility and achieved sample comparisons for specific modules
• Table 16–17: Data discrepancies and editing actions
• Table 18–19: Weighting factor calculations and final weights
• Table 21: Internet use estimates (2006 and 2014)
• Table 22–24: Profiles and confidence intervals for final and non-respondent samples

Conclusion

The CCMS consumer survey was a comprehensive effort to understand UK credit card usage and behavior. It incorporated a modular design, robust sampling methods, and careful data weighting to ensure national representativeness. The survey was conducted in two stages, with adjustments made to module allocation rules based on early findings. The final dataset included 39,837 valid interviews, with data cleaning and weighting processes ensuring reliability and accuracy for the FCA's analysis.