With market research companies being voted 6th in a 2018 Global Business Research Network (GBRN) trust survey across 11 different business sectors, it is imperative that market researchers ensure that their research is robust to gain the trust of their clients and participants. Robust research eliminates bias. Bias is unfair and makes research inaccurate and less trustworthy. It is difficult to completely remove bias as there are many variables that can influence data, however, bias can be dramatically reduced if research is planned correctly; this starts on platform and by following these 6 key steps you will help to reduce bias from your research on your insight platform:
1. Plan your Strategy Carefully from the Start
Strategies to reduce bias need to be considered from the beginning when establishing your platform. Think about your audience. Firstly, ensure your customers are able to easily access your platform and that they easily understand what they are required to do regardless of whether they are accessing via a link from a desktop or mobile or through an app; making things complicated from the start will put off even the keenest respondent.
 |
Tweet This |
|---|---|
| Strategies to reduce bias should be planned and implemented from the very beginning of the research experience, as bias can occur at any time through the very smallest of actions. |
Think carefully about the type of platform you require for your research, should it be branded or neutral? Consumer branded platforms will attract more brand loyalists as community members but if you wish to target your current customer base then this could suit better than an unbranded/neutral platform.
2. Ensure the Panel on your Platform are a Representative Sample of your Customer Base
Always ensure your customer base is large enough on your platform to gain representative samples from the population for any data you wish to collect. Make sure you have a mix of customers and non-customers to reduce sample-bias.
Establish representative samples of your target population for any research to make sure your target demographic are well represented. If you include or exclude specific respondents then you may get skewed results. Any data collected from a small sample or from a non-representative customer base will also bias your results.
3. Choose the Correct Tools to Gain Results and Think Carefully about the Set-up of your Platform
Surveys are not always the most appropriate form of market research data collection. Having a platform that supports a range of different tools and techniques for data collection will allow you to hone your questions and target your audience better to gain informed responses. Using a variety of methodologies, e.g. qualitative and quantitative methods to triangulate results, will help to give a more complete picture of the results during analysis.
Each methodology should be set up correctly on your platform to reduce bias, for example when asking a poll question or asking participants to respond to an image, such as those in a FlexMR’s SmartboardMR, ensure participants cannot see any results, or other participants’ comments until after they have voted/made their own comments to avoid any respondent bias. Or when setting up a survey, it may be more appropriate to randomise any answer options so participants do not all see answers in the same order so there is no selection bias.
4. Work with a Good Online Panel Provider who will Overcome any Limitations of Online Research
Online research has many benefits over traditional research methods, however, understanding the limits of your online platform during analysis will help you during the design phase of any methodologies as well as during interpretation of data. For example, if you understand that those who are able to participate in your online surveys are those who have access to the internet, or that members of online research panels may decide to answer questions based on what they think will yield more incentives, or in some cases try to answer questions more than once to gain further incentives, then you can put procedures in place to reduce this bias from the outset to ensure your data does not become skewed.
To overcome the bias work with a good online panel provider who will ensure that this does not occur, they will be able to prevent a respondent from joining the same survey more than once. If you do use multiple panel companies, make sure you ask the provider to disclose all the panel sources they are using as data quality may be compromised when a respondent is a member of two or more panels and receives the same survey more than once thus duplicating data.
5. Gain a Second Opinion to Reduce Researcher Bias
The way a research methodology is designed can have a big impact on your data. As mentioned above bias can be created by implementation of a singular research method but it can also be biased by having a single researcher on a project. Always seek a second or third opinion or consult with an online research partner who will have a team of researchers whose insight into platform or methodology design will reduce any subconscious bias. This is the same when interpreting results and making any recommendations.
|
Tweet This |
|---|---|
| While Respondent Bias is a danger to objective insights that needs to be addressed, so too is Researcher Bias, which can be just as harmful but trickier to identify. |
For example, focus groups are a great platform for group insight, bouncing ideas around and gaining context, however they can be biased (especially in a face-to-face situation) by overpowering or overtly shy personalities, online focus groups reduce much of this bias but one still has to ensure questions are not leading and that everyone in the group has a chance to respond and participate equally.
6. Reduce Response Bias on your Platform by Asking Simple and Logical Questions
Questions should be clear and simple and not double-barrelled or leading. Questions that are confusing or unclear can cause respondents to provide inaccurate answers. This can lead to inaccurate conclusions and therefore incorrect insights. Make sure your survey questions are not open to misinterpretation, especially if any of your respondents are not native speakers.
Make sure your questions on your insight platform in any methodology make logical sense. It may be more appropriate to place personal questions towards the end of a survey for instance to reduce drop-out rates of specific groups or you may need to ask about current practise or use of an item before you suggest alternatives. Keeping time periods short and relevant can also reduce the response bias for example, if you ask a respondent about something they encountered during the previous month as opposed to the previous year they are more likely to remember and respond accurately. Additionally, answer options should also be concise and precise to avoid misinterpretation and response bias.
Others ways to reduce response bias is to avoid asking respondents to use their own knowledge or experience to answer questions especially in comparisons for example compare this product to the one you currently use unless everyone uses the same products and therefore accurate comparisons can be made. If acronyms or jargon are to be used (although try to avoid this if at all possible to reduce confusion) then ensure you include details to explain the meaning of these in the questions.
Online research companies will assist in constructing a well-structured survey on an insight platform and can help to remove any incomplete questions or those who speed or flat line through a survey not taking time to answer thoughtfully, by using specific metrics during the fieldwork to remove response bias.
Eliminating Bias for Good
In conclusion, plan carefully from the start. Take the time to think about your customer base and ensure you have a representative sample from your population. Consider triangulation of methodologies to in order to eliminate as much bias as possible and gain second opinions to reduce researcher and response bias. An online research partner will be able to provide a multifaceted approach to design, analysis and insight for your research platform to reduce bias from the outset and throughout your research programme.
