Cognitive biases

Conquering Data Bias

Konstantinos Kattidis

Data Analytics Lead

Beyond data collection bias

Person making choices

Data bias can manifest in during data analysis, model development and interpretation
Cognitive biases are systematic patterns of deviation from norm or rationality in judgment

The tendency to seek, interpret, and remember information in a way that confirms pre-existing beliefs or hypotheses

This distorts the interpretation of results and hinders the discovery of novel insights
For instance:
- A marketing manager analyzing customer feedback about a new product
- Overlooking negative feedback that challenges their belief in its success

Marketing manager ignoring negative feedback

The tendency to overestimate one's own abilities, knowledge, or judgments

This bias can lead analysts to be overly confident in the accuracy or reliability of their findings

Person confident on analysis results

The tendency to give more weight or importance to recent events or information when making judgments or decisions

It can lead analysts to prioritize recent data points or trends over historical data
Potentially overlooking long-term patterns or trends that may be relevant to the analysis

Line graph with focus on recent data points

Sculpting the past to fit the present

The ways in which our memories can be distorted or influenced by various factors

A sales manager reviewing past sales performance
Remembering successful outcomes while downplaying or forgetting instances where similar tactics failed

Person recalling

Influence of easily accessible information

A mental shortcut that relies on immediate examples that come to a person's mind when making judgments or decisions

For example:
- A risk analyst overestimating the likelihood of a data breach
- Because recent news headlines have highlighted similar incidents

Analyst thinking about recent events influencing his analysis

Influence of initial reference points

Occurs when individuals rely too heavily on initial information, when making decisions or estimations

For example:
- An analyst starts with an initial estimate of a parameter
- Unconsciously adjusting their subsequent estimates or interpretations around that initial anchor

Person looking at a report with an anchor

Cognitive biases often lead to the reinforcement of existing beliefs and perspectives

Feedback loops

This often creates a feedback loop, where individuals consistently encounter information that aligns with their views
It further solidifies their beliefs over time
It leads to closed-mindedness and a resistance to considering alternative perspectives

Conquering Data Bias