Our team of research, data and psychology specialists will help you understand your data needs and provide bespoke insight and solutions.
RepGraph provides statistical analysis and interpretation of research data. This may be research carried out by RepGraph, commissioned by ourselves on your behalf or data gathered by your organisation. We also conduct data cleansing exercises on a bespoke basis. RepGraph develops domain knowledge through research and communication with our clients. This not only assists with data validation but also guarantees that outputs such as online reports and presentations have the right context and are relevant to the target audience.
In the early 1970s, Amos Tversky and Daniel Kahneman introduced the term ‘cognitive bias’ to describe people’s systematic but purportedly flawed patterns of responses to judgment and decision problems. Cognitive biases are just tools, useful in the right contexts, harmful in others. They influence how we think about and interpret data and can sometimes lead to errors in judgement and flawed or subjective decision-making.
During World War II, some bomber aircraft were shot down over Germany whilst others would return with bullet holes and severe damage. Allied researchers mapped the areas that were most commonly hit by enemy fire and the data showed a clear pattern. In the image you can see that most damage is to the wings and body of the plane but not the cockpit, engine, and parts of the tail. The analysts concluded that the damaged areas needed to be strengthened.
The American military asked mathematician Abraham Wald to study how best to protect aircraft from being shot down. Wald pointed out that the data had a critical flaw – it did not include those bombers that did not return. What the researchers had created was a map of the areas that the bomber could be shot and still survive. This insight led to the armour being reinforced on the parts of the plane where there were no bullet holes.
Human insight is key – while amassing data is critical, missing data may be more important to solving a problem than the data you have in front of you.
This logical error is known as ‘survivorship bias’ when we assume that success tells the whole story and when we don’t adequately account for failure.
HOW WE DO IT
Identify and plan the best approach for your business
Apply qualitative and quantitative research methods to provide unique insights
Provide outcomes in a clear, digestible and most of all practical format
Research DesignRepGraph employs a client-requirements driven approach to research design. Our psychology expertise uses cognitive bias to formulate questions and challenge research outcomes.
We use an iterative methodology in project planning and undertake small scale proof of concept exercises to inform project decision making and assist in refining requirements.
Data cleansingData cannot be properly analysed unless it is as clean and accurate as possible. Analysis of uncleansed data can lead to inaccurate results, incorrect interpretation and even false outcomes.
RepGraph uses a combination of techniques to cleanse and organise data including grouping and deduplication, natural language processing, automated validation and cleansing metrics.
Data ScienceData sets in the modern world have become large, complex and difficult to manage using traditional tools. The RepGraph team possesses seasoned expertise in modern data science tools including python data libraries for ingestion and processing, SPSS and R for in-depth analysis and statistical interpretation and Power BI and Tableau for reporting. RepGraph also possesses strong expertise in data extraction and transformation using industry-standard tools.
Client SurveysRepGraph has developed a wide variety of surveys and questionnaires as stand-alone projects or as part of large research exercises including:
- Client/Member sentiment surveys
- Research into best practice
- Annual state of the industry publications
- Public attitudes to industry surveys. We are expert in constructing questionnaires that ensure the wording is appropriate, clear, understood and unbiased