Evidence about a product’s effectiveness can help district leaders decide which educational technologies or instructional resources to purchase and use.
Mathematica Policy Research created a guide to help educators determine which claims can be trusted and which are less reliable. The following is a summary of evidence that educators are most likely to encounter, from weakest to strongest:
Weak: Anecdotal
Consists of personal descriptions or claims based on one person’s experience or subjective impressions.
Example: My students love using product X. They use it for about 20 minutes every day. On average, my first-grade class is working at a middle of second-grade level.
Average: Descriptive
Summarizes characteristics of participants and outcomes over a period of time without evidence. It is commonly found in marketing materials and news articles.
Example: An infographic displays positive results—without a comparison group.
Moderately strong: Correlational
Identifies the relationship between an educational condition or initiative (such as using an educational technology) and a specific outcome (such as math test scores). It’s a good starting point, but it cannot rule out other possible explanations for the differences in outcomes between users and nonusers.
Example: Middle school students participating in a blended-learning reading program showed increased gains in math skills—up to nearly 50 percent higher in some cases—over the national average.
Strongest: Causal
Usually conducted by independent evaluators and compares apples to apples by ensuring the only difference between the group that participated in the program and a comparison group is the program itself.
Example: Any independent evaluation that uses a randomized controlled trial design.
Source: “Understanding Types of Evidence: A Guide for Educators,” Mathematica Policy Research, 2016.