Consumer Contact prides itself on its reputation for providing high quality data collection and loyalty measurement and tracking services to our clients. When a company decides to make an investment into conducting a project with us, they expect accurate, actionable results. These results are very often the basis for strategic decision making within a company. A client should not have to worry that the quality of their data is compromised because of potential sample bias. Sample bias occurs during the survey process when some segments of the population being studied are either over- or under- represented in the sample. Sometimes this happens because 1000 people answer a survey one month, compared to 2000 the next, or too many people over the age of 50 respond to a survey compared too few responses from those under 30. If left uncontrolled, these variations can adversely affect the accuracy of your resulting statistics and the conclusions you draw from them.  Your sample composition needs to be consistent across time periods and proportionately representative of your target market. When this isn’t possible, Consumer Contact’s data weighting practices should be employed to improve sample proportionality and consistency and help to reduce the effects of sample bias.

Sample Bias

Complex data collection projects often require collecting feedback from several different channels within a business or from several different demographics across the country. It is difficult to predict who will choose to provide their feedback and who will not, and levels of cooperation can vary from one month to the next.

Examples of situations where sample bias may arise include:

• Older people are often more willing to complete a survey than younger people leading to an under-representation of younger opinions.
• Sometimes males (or females) may be more likely to complete a survey. You want your data to reflect the actual gender split of your target market.
• Data needs to be comparable from one month to the next or from one area of a business to another. Bias is found when different amounts of customers respond to ongoing surveys each month or different areas of a company hold different weight based on the number of high-value clients they have. Adding or subtracting weight in these situations can improve the accuracy of your data.

Why Weight

You should never be afraid to weight data. When people adjust data they often think they are manipulating results, which is not true. Adjusting for sample bias actually has no true impact on the original data collected; it simply makes the results more representative of the target market studied, allowing companies to focus with more detail on what areas of business can be improved. When deciding if your sample is proportionate or not, one should always weight based off of the best known data source. For a general population survey, census data is the most accurate source to reference weighting ideals on demographics and population distribution. If you are surveying banking clients then you should reference the data available from your customer bank lists, and so on.

Weighting Example:

Question 1: Do you agree with the statement: Valentine’s Day is an important holiday?

Results:  50%- yes   50%- no

Breakdown of YES answers:

Before Weighting                                                                                     After Weighting

The example above is a simple breakdown of how weighting works. For this study, more women were interviewed than men; 600 versus 400. This could be for no other reason that more women were available to complete the study during the time it was being conducted. Because more women responded, their opinions are represented more than the men, resulting in a figure that states 54% of the Canadian population thinks Valentine’s Day is an important holiday. In reality, the figure is skewed to more heavily represent female opinions than male opinions since more women were surveyed. The figure does not accurately reflect what a combined population of both men and women think; it’s more biased to what women think.

In order to accurately represent the Canadian population which is roughly 50% men and 50% women, we weighed the data to represent an even 500 responses from both sexes, giving men a higher weight (1.25) since they initially provided fewer responses and women a lower weight  (0.834) since they had more responses. A computer works out the math to ensure the weighting formula is accurate. After weighting the data, the amount of men and women who agreed with Question 1 remains at 30% and 70% respectively. The only statistic that changes is total amount of people who chose YES: decreasing from 54% to 50%, representing men and women equally. The accuracy of the data was not harmed by the weighting, rather it was improved.

The Foundation for Quality Results

At the core of our CATI phone surveying process sit our specially selected qualified interviewers. We employ roughly 1,200 interviewers across our five call centres and 470 CATI stations.  A strong and knowledgeable interviewer can attract more respondents, provide knowledgeable, detailed answers to questions and ensure an enjoyable and engaging phone survey experience. Though we require a comprehensive range of skills from all interviewing staff, we understand the importance of highly trained Business-to-Business (B2B) interviewers and how they represent your company. At Consumer Contact we remove the stress out of B2B telephone research by creating a team of Executive Research Interviewers.

B2B surveys can contain intricate and detailed questions directed towards high-value clients and business executives. Potential Executive Research Interviewers undergo rigorous training and demonstrate a list of required skills before achieving their ERI designation.

CCLM Executive Research Interviewers are:

• Able to reach the right respondents and gather the right information
• Experienced interviewers who have been identified as top performers
  • selected from the top 10% of performers
  • Individually selected by Business-to-Business clients
  • Briefed personally by clients on the details of their business and on any essential background information
  • Veteran professionals with more than 2 years of tenure with Consumer Contact
  • Knowledgeable in the language of your business
    • E.g. Finance, Information technology, health care
    • Selected for their business acumen and professional demeanor
      • Conversational and confident
      • Given extensive B2B interview training
      • Skilled in negotiating corporate barriers and “gatekeepers” to contact senior executives
      • Briefed extensively on company background and research objectives
      • Able to recruit executives for online research

Executive Research Interviewers are self-motivated and possess perseverance in maintaining quality standards while also ensuring that interviews are conducted in a language that respondents understand. While most interviews are completed in English, French and Spanish, projects have also been executed in Cantonese, Mandarin, Punjabi, Italian and Portuguese.  Whatever your marketing or consulting objectives, at Consumer Contact, we are dedicated to meeting the needs of our clients, and that begins with a great team of employees committed to delivering the industry’s most professional service, most accurate information and most valuable and actionable results.