Pre-experimental Designs in Educational Research: An Overview

When educational researchers want to test whether a new teaching method, curriculum, or intervention actually works, they need a structured way to collect and interpret data. That’s where research designs come in. Not all research settings allow for the gold standard of random assignment and control groups – and in those cases, researchers often turn to pre-experimental designs. These are the most basic forms of experimental research, characterized by their simplicity and by the limitations they carry. Understanding these designs – what they are, how they work, and where they fall short – is essential for anyone evaluating or conducting educational research.

What are pre-experimental designs?
Key characteristics of pre-experimental designs
The three main types of pre-experimental designs
1. The one-shot case study design
2. The one group pre-test post-test design
3. The static group comparison design
How pre-experimental designs fit into the larger research framework
Practical implications for educational researchers

What are pre-experimental designs?

Pre-experimental designs are research frameworks in which a subject or group is observed after a treatment or intervention has been applied, in order to test whether that treatment has the potential to cause change. As the SAGE Encyclopedia of Educational Research explains, the prefix “pre-” conveys two meanings: these designs are more rudimentary relative to true experiments, and they often serve as preparatory exploration before a full experiment is undertaken.

According to the SAGE Encyclopedia of Research Design, pre-experimental designs either fail to include a pretest, a control or comparison group, or both – and crucially, no randomization procedures are used to control for extraneous variables. This is precisely what makes them “pre-experimental”: they follow some basic steps of experimentation, but do not achieve the level of control required for valid causal inferences.

That said, they are not without value. These designs are cost-effective, practical, and particularly useful when researchers face constraints of time, resources, ethics, or participant availability. They are best understood as a first step – a way to test whether an intervention shows enough promise to justify a more rigorous study.

Key characteristics of pre-experimental designs

Before diving into the specific types, it helps to understand the shared features that define this category of designs:

No random assignment: Participants are not randomly allocated to groups. This means the researcher cannot rule out pre-existing differences between participants that might influence outcomes independently of the intervention.

Limited or no control group: Most pre-experimental designs either lack a control group entirely or include one without proper equivalence checks. Without a control group, there is no reliable baseline for comparison.

Weak internal validity: Because extraneous variables are not controlled, it is difficult to conclude that observed changes are actually caused by the intervention. As the Distance Learning Institute notes, this makes it hard to rule out alternative explanations for observed effects.

Exploratory in nature: These designs are suited for generating hypotheses, gathering preliminary data, or testing the feasibility of a study – not for drawing definitive causal conclusions.

The three main types of pre-experimental designs

As first systematized by psychologists Donald Campbell and Julian Stanley in their landmark 1963 monograph, three designs are considered standard pre-experiments in educational research: the One-Shot Case Study, the One Group Pre-test Post-test Design, and the Static Group Comparison Design.

1. The one-shot case study design

This is the simplest and most minimal of all pre-experimental designs. A single group of participants receives a treatment or intervention, and their outcomes are measured just once – after the treatment. There is no pretest and no control group. In research notation, it is often represented as: X → O (treatment, then observation).

For example, a teacher introduces a new problem-solving technique to a class, and then assesses student performance at the end of the week. The score obtained is the only data point available.

The core problem with this design is that without a pretest or control group, there is no baseline against which to measure change. As researchers note, the design is highly vulnerable to validity threats, including history effects (external events between intervention and observation), maturation effects (natural changes in participants over time), and confounding variables. Any observed result cannot be conclusively attributed to the intervention alone.

Despite its weaknesses, the one-shot case study still has a place in exploratory work. It can be useful for pilot testing an intervention or when a pretest is simply impossible to administer. Researchers must, however, interpret results with significant caution and avoid making causal claims.

2. The one group pre-test post-test design

This design adds a pretest to the one-shot case study, making it marginally stronger. Here, the same group is measured before the intervention (O₁), receives the treatment (X), and is then measured again after (O₂). The difference between O₁ and O₂ is used to infer whether the treatment had an effect. The notation is: O₁ → X → O₂.

For instance, a researcher assessing the impact of a reading intervention program measures students’ reading levels at the start of the semester (pretest), implements the program, then measures reading levels again at the end (posttest). If scores improve from O₁ to O₂, the intervention appears effective.

The addition of the pretest is a meaningful improvement – it provides a baseline and allows researchers to document change over time. As the University of Texas at Arlington’s research methods resource explains, researchers may be able to claim that participants experienced change in the dependent variable, but cannot attribute that change to the treatment without a comparison group.

The design remains susceptible to several threats to internal validity. History is a major concern – other events happening between the pretest and posttest (a school event, external media, seasonal changes) could explain the improvement. Maturation is another: students naturally develop over time, and any improvement may reflect growth rather than the intervention. The testing effect is also a risk – students who have already taken the pretest may perform better on the posttest simply because they are familiar with the format, not because the intervention worked. As research methodology texts consistently emphasize, these alternative explanations mean that conclusions from this design must remain tentative.

3. The static group comparison design

The static group comparison design introduces a second group into the picture, making it the most complex of the three pre-experimental designs. Two groups are compared: one that receives the treatment (experimental group) and one that does not (comparison group). Both groups are measured after the treatment only – there is no pretest for either group. The notation is: X O₁ / O₂ (treatment group measured, comparison group measured).

Consider a researcher studying whether a new digital learning tool improves mathematics performance. One class uses the tool for a month (experimental group), while another class continues with the traditional method (comparison group). At the end of the month, both classes take the same test, and the scores are compared.

Compared to the previous two designs, the static group comparison offers the advantage of a reference point – the comparison group. However, according to the SAGE Encyclopedia of Educational Research, the design is vulnerable to three key threats to internal validity: selection bias (the groups may have been systematically different before the intervention), mortality (differential dropout rates between groups may affect posttest results), and maturation. Without random assignment, there is no way to confirm the groups were equivalent at the start.

The design is particularly applicable in situations where random assignment is ethically or practically impossible. As Quantifying Health notes, it can also offer reasonable external validity since it often uses pre-existing groups in real-world settings, and findings may generalize to similar populations. The results, however, should be treated as suggestive rather than conclusive evidence of a treatment effect.

How pre-experimental designs fit into the larger research framework

It is worth placing these designs in their proper context. The Office of Justice Programs summarizes it clearly: pre-experimental designs are used to draw preliminary inferences about the effect of an intervention, but the absence of adequate controls makes definitive scientific conclusions impossible. They sit below quasi-experimental and true experimental designs on the hierarchy of research rigor.

True experimental designs – which require random assignment, control groups, and both pre- and post-tests – are the standard for establishing causality. Quasi-experimental designs occupy a middle ground, lacking random assignment but including comparison groups. Pre-experimental designs, by contrast, are the starting point – valuable for exploration and hypothesis generation, but not for establishing firm cause-and-effect relationships.

Researchers in education and social science frequently encounter these designs in program evaluations, pilot studies, and field-based investigations where the conditions for a true experiment simply do not exist. As doctoral-level social work research resources note, pre-experimental designs are useful when researchers are developing new interventions, testing new measurement instruments, or building toward more rigorous future designs. A negative or null result from a pre-experimental study is itself useful – it may prompt a rethinking of whether an intervention is worth pursuing at all.

Practical implications for educational researchers

Educators and researchers who encounter or use pre-experimental designs should keep a few guiding principles in mind. First, be transparent about the design’s limitations when reporting findings. Readers and policymakers need to understand that pre-experimental results cannot support strong causal claims. Second, treat results as preliminary. A promising outcome from a one-group pre-test post-test study is a reason to conduct a more rigorous follow-up, not a reason to implement a program district-wide. Third, where possible, plan ahead. Even if random assignment is not feasible, incorporating a comparison group or pretest measurement significantly strengthens a study’s credibility.

The value of pre-experimental designs ultimately lies in their accessibility. They make it possible to gather data under real-world constraints, and when interpreted carefully, they can meaningfully contribute to the evidence base for educational practice. The key is knowing exactly what these designs can – and cannot – tell you.

What do you think? If you were evaluating a new classroom intervention at your school and could not randomly assign students to groups, which of the three pre-experimental designs would you choose and why? And how would you account for the limitations of that design when presenting your findings to school administrators?

References

PDF 📄

Table of Contents

What are pre-experimental designs? 🔗

Key characteristics of pre-experimental designs 🔗

The three main types of pre-experimental designs 🔗

1. The one-shot case study design 🔗

2. The one group pre-test post-test design 🔗

3. The static group comparison design 🔗

How pre-experimental designs fit into the larger research framework 🔗

Practical implications for educational researchers 🔗