What is Correlational Research?

Correlational research is a type of research that seeks to establish relationships between variables without manipulating them directly. In edtech product research, a correlational study can help a company examine relationships between their products' usage and learning outcomes, engagement levels, and other user experiences. A company can utilize these insights to inform product development decisions.

Correlational studies use statistical analysis of quantitative data to identify connections. In correlational research, no cause and effect relationship is assumed or established, but rather the focus is on measuring the strength and direction of the relationship. Compared to other types of research, correlational research tends to require larger data sets in order to identify relationships. For example, correlational research on a math tool may examine standardized testing scores from across three high schools in two school districts.

When longer-term data is available to researchers (say a reading application has been used for several years by fifth graders in one school district) along with control data (only half of the district’s fifth grade classrooms used the app), quasi-experimental techniques may also be used. Quasi-experimental studies aim to provide insights into potential cause and effect relationships, this time between an independent and dependent variable—meaning one variable is manipulated and the other variable is affected. While quasi-experimental methods provide more control over variables, correlational research is often more feasible and cost-effective for edtech companies.

Depending on the edtech tool being studied and the company’s research goals, a correlational study is codesigned to examine outcomes following the implementation and consistent use of said product. One example of a correlational study conducted by Leanlab is a correlational and quasi-experimental study with SpeakAgent, an academic language learning platform for K-12 public schools. Leveraging a district-wide partnership with data from nearly 1,600 middle school classrooms, our data analysis explored the effect of using Speak Agent on standardized test scores.

LEARN ABOUT LEANLAB’S CODESIGN PRODUCT DEVELOPMENT AND EFFICACY RESEARCH SERVICES FOR EDUCATION COMPANIES

Who are Correlational Studies for?

Correlational studies are for edtech companies looking to understand the relationship between the use of their product and learning outcomes. The intensive nature of correlational research—and the scope and quantity of data it requires—means that correlational studies are best for later-stage companies with an established user base. While there are no specific industry-wide usage minimums, at Leanlab we recommend that products have been used in a school or learning setting for at least a year, and used with fidelity previously established through implementation research.

Correlational studies are valuable for companies looking to secure ESSA certification for their product, as positive significant findings can demonstrate promising evidence (ESSA Tier III) to schools seeking edtech solutions. Correlational studies can also be a useful step before undertaking more rigorous efficacy research using specific quasi-experimental or randomized controlled trial methods that may be eligible for ESSA II and I Tiers of Evidence.

Why Conduct Correlational Research?

COllect preliminary evidence of impact

Edtech companies develop products in order to improve student outcomes, so logically companies are eager to conduct research that demonstrates a correlation between product usage and a specific outcome. Working with school partners that provide ongoing data collection and third party researchers to objectively analyze the data, correlational research can generate preliminary evidence of impact.

Preliminary evidence of impact can support product led growth for companies as a compelling factor for schools in the procurement process. For companies seeking evidence to qualify for ESSA Tiers of Evidence, correlational research that meets certain sample criteria and finds at least one statistically significant positive outcome may qualify for ESSA III. When quasi-experimental methods are used and study guidelines are met, moderate evidence may be found that qualifies for ESSA II.

Guiding Questions for Correlational Research

Research questions for correlational studies are developed with the product and intended outcome in mind, though key guiding questions are at the core of all correlational studies.

What is the relationship between use of the product and intended student or teacher outcomes?

This initial guiding question explores the relationship between the use of a product and the intended outcomes. When students begin using a new tool, do test results improve? Did the product impact the intended outcomes?

Is there a correlation between use of product and improved outcomes?

Demonstrated correlation between improved student outcomes and the use of a company’s product is compelling for companies, and as such is a guiding question for correlational research. Does use of your daily Spanish learning app correlate to improved student comprehension? Does use of your company’s learning management system’s homework calendar correlate with improved student assignment completion? These types of questions are explored through statistical analysis.

For quasi-experimental studies: What is the effect of the product on intended outcomes?

A step further than correlational research, quasi-experimental studies ask what effect the product has on intended outcomes. With quasi-experimental studies, a company can begin to gather insights into potential cause and effect relationships between the use of their product and student outcomes. Quasi-experimental studies can set a company up for conducting a subsequent randomized controlled trial (RCT) study and potential ESSA Tier I certification with positive statistically significant results. RCT studies are the gold standard in research for establishing cause and effect relationships.

How to Prepare for Correlational Research

Establish foundational evidence

Companies eager to establish evidence of impact for their products may want to jump straight to correlational research. However, before student or teacher outcomes can be measured in a correlational study, we must ensure that the tool is being used as intended and with high user engagement. This foundation of evidence is established through usability, feasibility, and implementation studies. Conducting formative research also helps companies and researchers establish a relationship with long-term school partners, allowing for time-series data that can be used in future studies.

The very basic building blocks of the product’s research journey should also include a sound logic model—laying out how people should interact with the product to achieve the intended outcomes—interoperability, and an established user base that includes student use data and outcome data.

Correlational research is a valuable part of the research journey. It can uncover positive efficacy findings and offer evidence of impact to schools, or provide critical insights to inform future product development. It can lay the groundwork for larger studies including randomized controlled trials that further demonstrate the value of a product to teachers or students.