A retail company wants to follow responsible practices when collecting data to build a product recommendation model. Which data-collection practice helps reduce bias?

Ensuring the dataset is balanced and drawn from a diverse set of customers is crucial for reducing bias. A balanced and diverse dataset represents various customer segments accurately, preventing the model from disproportionately favoring or underrepresenting certain groups. Using data only from customers matching the company's overall demographic profile can introduce bias if that profile isn't representative of the entire customer base or target market. Collecting data only from customers with a prior purchase history creates a biased dataset that excludes potential new customers or those with different purchasing patterns. Using data sourced only from a publicly available dataset doesn't guarantee diversity or balance and may not be relevant to the company's specific customer base, potentially introducing external biases.