Video tool

Type of tool: Workshop

Phase: Deployment

COLLABORATIVE CORRELATION DATA ANALYSIS

Include citizens’ perspective in data analysis

People involved:

-40

Duration:

1 hour

Author:

ISGlobal, Ideas for Change

The Challenge

In the analysis of epidemiological data, researchers often look for correlation among two or more variables. For instance, when the values of air pollution rise, human attention decreases. However, it is more complex than that! Researchers know very well that different factors, other than the two variables under study, can influence a correlation. For instance, our attention might decrease because we slept badly the previous night. Thus, the analysis of data must include these influencing factors in order to ensure that the correlations we observe are not influenced by those factors. But, what are relevant factors that should be included in the analysis of correlations?   

Involving citizens in this phase allows researchers to include new viewpoints on the analysis, taking advantage of the situated knowledge of affected citizens who can provide their first-hand experience of the problem and shed light on new scenarios. 

How can we include citizens’ experience of the problem at stake in the data analysis processes?

The Tool

This tool allows you to foster citizens’ debates around different aspects of data analysis. Specifically, through this tool you will be able to identify the variables to be included in the correlation data analysis, according to the experience and perspective of citizens.

Discover the tool in action!

Read the case study and understand how this tool has been used in a real citizen science project.

CitieS-Health Collaborative Correlation Data Analysis

CitieS-Health Barcelona Pilot

What

The tool was used during a workshop attended by citizens who took part in the data collection of the CitieS-Health Barcelona pilot. Before the workshop, citizens received a report with a rough analysis of their data. The data were about individual exposures to air and noise pollution, as well as their daily responses to the survey that collect information about sleep quality, attention, stress and daily habits. The goal of the workshop was 1) to discuss the type of data analysis researchers planned to perform, 2) to provide an overview of all variables created from the data collected, 3) to identify with participants the variables they think can play an important influence in the correlation investigated and 4) to raise other research questions that could enrich the data analysis.

Why

To identify confounding variables to be included in correlation analysis.
To identify other research questions that could help understand and enrich the analysis of the data collected.
To leverage the first-hand experience of citizens of the problem at stake.

How
Provide an overview of the data collected We start by remembering participants about all the data collected. On a big sheet, you can provide an overview of the categories of the data which are available for the analysis (e.g. if you are conducting an epidemiological study, you could have the categories “Exposure outcomes”, “Health outcomes” and “Sociodemographic characteristics”. This puts everyone in the same position, allowing them to take part in the conversations that follow.
Co-creation activity 1: prioritize variables Among all the variables collected, you present participants the potential confounding variables that can have an effect on the correlation under investigation. By using sticky dots on a poster, or through an online tool such as Sli.do, each individual is asked to select the variables that, in their opinion, could influence the dependent variable more (in our case, our dependent variables were attention, sleep quality,stress and mood). Then, they are asked to sort the variable in order of importance. The workshop facilitator asks participants to share their thoughts with the rest of the group.
Co-creation activity 2: secondary questions This activity aims to collect ideas about other possible research questions that can be answered through the data collected. Participants work in groups and brainstorm possible questions taking into account all the data categories presented before. To facilitate the brainstorming, participants are provided with a poster or an online slide showing an overview of the categories of data collected.
CLOSE CASE STUDY
CONTINUE READING