In the data science world, you hear the terms descriptive analytics, predictive analytics, prescriptive analytics, or diagnostic analytics.

By analyzing historical and current data, predictive analytics anticipates events in the future.

Models consisting of multiple levels are also referred to as linear mixed, hierarchical, or mixed-effect models.

In computer science, artificial intelligence (AI) is the development of intelligent machines that mimic the capabilities of humans.

Difference-in-Difference is one of the most popular quasi-experimental techniques, backed by a large body of historical evidence.

Randomised controlled trial (RCT) is a popular evaluation design.

Regression Discontinuity Design is a quasi-experimental design and another way to deal with selection bias.

Change in government policy is brought about by influencing the opinions of leaders and decision-makers through policy advocacy.

It is often not possible to do a randomized controlled trial or RCT in a real-world setting for impact evaluation. 

As the name implies, summative evaluations provide a summative assessment of the programme.

Propensity Score Matching (PSM) has been used since the 1980s to reduce selection bias and produce better causal estimates.

Monitoring, evaluation, and research are critical learning components because they tell implementors about course corrections based on evidence-based decision-making.