Predicting data saturation in qualitative surveys with mathematical models from ecological research

Area of Science:

• Quantitative research methodologies
• Survey research design
• Statistical modeling in social sciences

Background:

• Data saturation is crucial for sample size determination in qualitative surveys.
• Current methods rely on researcher judgment, introducing subjectivity.
• Predicting saturation is challenging due to incomplete data during collection.

Purpose of the Study:

• To introduce mathematical modeling for predicting data saturation in surveys.
• To provide an objective method for determining optimal sample size.
• To extrapolate theme accumulation and guide data collection cessation.

Main Methods:

• Utilized a mathematical model based on a latent distribution of theme elicitation.
• Employed a mixture of zero-truncated binomial distributions to model theme accumulation.
• Validated the model using Monte Carlo simulations and real-world survey data on treatment burden.

Main Results:

• The model accurately predicted the number of themes with low estimation error (<3%).
• A stopping criterion based on the slope of the theme accumulation curve was effective.
• This criterion identified 97.5% of themes while minimizing redundant data collection.

Conclusions:

• Mathematical models, adapted from ecological research, offer accurate prediction of data saturation.
• This approach provides an objective and efficient method for sample size determination in qualitative surveys.
• The proposed method enhances research efficiency and reliability in theme saturation.