May 17, 2023

Four Tips for Avoiding Bias in Social Listening

Date & Time (GMT):
May 17, 2023 12:47 PM
Date & Time (EST):
May 17, 2023 12:47 PM

The Social Intelligence Lab community member, James Tattersfield, MD at Polar Insight gives his thinking on how to reduce bias in social listening projects.

When conducting any kind of research, it is virtually impossible to separate yourself completely from the data you want to collect. We all have known and unknown biases and these can impact the information that is collected, how it is analysed and the interpretations we provide.

Some of the most commonly observed biases include:

  • Selection bias
  • Self-selection bias
  • Recall bias
  • Observer bias
  • Survivorship bias
  • Omitted variable bias
  • Cause-effect bias
  • Funding bias
  • Cognitive bias

Before moving on, I highly recommend a quick Google search of the different biases outlined above. There’s a wealth of information out there worth investigating and it will transform how you think about gathering and interpreting information.

Coming from a research background me and my team are always thinking about how we can remove bias from our research to deliver better, objective findings for our clients.

To get you started, we've pulled together a list of approaches you can use to avoid bias and to make your research as objective as possible.

1. Use Multiple and Different Data Sources

Firstly, the use of multiple data sources - otherwise known as ‘triangulation’ - is a common method used to support the interpretation of data. By linking several data sources together you can compare and contrast, find similarities and build confidence that your findings are legitimate.

For a social intelligence project, this might be using existing in-house research findings on the topic. You could use multiple tools, like Brandwatch with Audiense or something like Mentionmapp. Or maybe a trends based tools like Google Trends to see how social mentions compared to search activity.

2. Use Multiple and Different People to Interpret the Data

Secondly, we recommend that you engage several people in interpreting your data. Consistency between interpretations improves the likelihood your findings are a fair representation of what is happening.

Here it makes sense to get people involved from as many different departments as possible.

As with all other activities at work, there's always underlying organisational needs as to why something like social listening might be being conducted. If possible, these should be mapped as they could help surface new insights from the data.

People need to look beyond their departmental needs / KPI to the wider context and objectives of their organisations - why does it exist, what is it trying to achieve, what kind of world does it want to see exist?

3. Examine alternative explanations

Always consider whether there are alternative explanations for your data. If you are careful to rule out or account for as many of these alternative explanations as possible, any interpretations will be more robust.

4. Review your findings and conclusions with others

Finally, ask others to review your conclusions. If you can, share your findings with people outside of your immediate peer group or department. A fresh pair of eyes can be hugely valuable.

If you like to use any other methodologies to reduce bias in your social intelligence work, we’d love to hear about them.

Or view the interview on LinkedIn

This interview was recorded via LinkedIn Live, if you prefer to view on LinkedIn, click the button below.

View Interview

See related content

Webinar & Panel
How to extract insights from social data
Webinar & Panel
How live insights can prevent your brand from becoming an unintended target in a year of global elections
Webinar & Panel
See, hear, and feel the social pulse: deeper insights on social listening with Gen AI