Free scoping read. A written assessment of your manuscript in two business days, at no cost.

Start one
Contents

Academy · Research Data Analytics & AI · Pre-experimental designs

Types of Pre-Experimental Research Design

Pre-experimental designs include an intervention and a measurement but no control over who receives the intervention. They are quick, inexpensive and useful for piloting, but they cannot establish that the intervention caused the observed change, which is the claim authors most often make from them.

Pubrica AcademyWritten by the medical writing and biostatistics teams Reviewed byPubrica biostatistics team Published2 October 2026 Last checked2 October 2026 Reading time4 minutes

Defining features of a pre-experimental design

A pre-experimental design applies an intervention to a group and measures an outcome without random allocation and usually without a control group. It ranks below quasi-experimental designs, which lack randomisation but retain a comparison group, and well below true experimental designs, which have both.

The defining feature is not the absence of statistics but the absence of control. Nothing in the design rules out the alternative explanations for a change: maturation, history, regression to the mean, testing effects, or simply that the people who received the intervention differed from those who did not.

What these designs can and cannot do

They can establish that a change occurred, generate hypotheses, test the feasibility of a protocol and inform the sample size of a later study. They cannot establish causation. A paper that reports a pre-experimental result and concludes that the intervention “improved” the outcome has claimed more than the design permits, and reviewers will identify this.

The three pre-experimental designs

Three designs account for almost all pre-experimental work, and they differ in the extent to which they can exclude alternative explanations.

One-shot case study. A single group receives the intervention and is measured once, afterwards. Written as X → O.

  • Use it to describe what happened after an intervention when no baseline exists.
  • The weakness — with no baseline measurement and no comparison group, the result has no point of reference. It describes a state, not an effect.

One-group pretest–posttest. One group is measured, receives the intervention, and is measured again. Written as O₁ → X → O₂.

  • Use it to measure change over the period of an intervention, and to pilot the outcome instrument.
  • The weakness — the change may be due to maturation, to events outside the study, to regression to the mean if the group was selected on an extreme score, or to the practice effect of having taken the test once already.

Static group comparison. Two existing groups are compared after one has received the intervention, with no baseline and no random allocation. Written as X → O₁ alongside O₂.

  • Use it to compare naturally occurring groups when allocation is impossible.
  • The weakness — selection bias. Because no mechanism ensured that the groups were comparable at the outset, any subsequent difference may simply reflect a pre-existing one.

Conduct of a pre-experimental study

The sequence is short, which partly explains the appeal of these designs, and each step constrains what the report may claim.

The sequence of a pre-experimental study, from defining the problem through to interpreting results against the design's limits.
The sequence of a pre-experimental study, from defining the problem through to interpreting results against the design's limits.

Sampling is the step that most compromises validity. A convenience sample is typical of this design and is the main reason the result does not generalise; it should therefore be stated in the limitations rather than left for a reviewer to identify.

Analysis of the data

Analysis is usually descriptive, with simple inferential tests where a before-and-after comparison exists.

  • One-shot case study — descriptive statistics only. There is no comparison to test.
  • One-group pretest–posttest — a paired t-test where the difference is approximately normal, or the Wilcoxon signed-rank test where it is not.
  • Static group comparison — an independent-samples t-test or the Mann–Whitney U test, with the caveat that the groups were never equivalent.

The effect size and its confidence interval should be reported alongside any p-value. In a design of such limited strength, the magnitude of the change and its precision are more informative than whether it crossed a significance threshold.

Advantages and their associated limitations

Where they earn their place

These designs are fast and inexpensive, and feasible where randomisation is impossible or unethical. They are useful for piloting an instrument, testing recruitment assumptions, and producing the variance estimate that a later study requires for its sample size calculation.

What they cannot support

These designs cannot support a causal claim and provide no defence against maturation, history, regression to the mean or testing effects. Selection bias is uncontrolled where two groups are compared. Findings rarely generalise beyond the sample. Because the design is weak, only replication can make a finding from it credible.

Such a study should be reported under STROBE where it is observational in character, the design should be named in the methods, and the uncontrolled threats should be stated explicitly in the limitations rather than in a generic sentence about sample size 2.

Key points for authors and researchers

  • These designs can demonstrate that a change occurred, but not that the intervention caused it.
  • The design should be named in the methods section, and the uncontrolled threats stated clearly in the limitations.
  • The effect size and its confidence interval should be reported; in a design of this weakness, they are more informative than a p-value.
  • These designs are best used as pilot studies to assess feasibility, instrument performance and the variance required for a later sample size calculation.

Frequently asked questions

Yes, provided that it is framed accurately: as a pilot, a feasibility study, or a descriptive account. Journals reject such studies when the discussion claims causation, not because of the inherent weakness of the design.

A quasi-experimental design keeps a comparison group but allocates without randomisation. A pre-experimental design usually has neither. The additional comparison group is what enables a quasi-experimental design to rule out some alternative explanations.

This is one of its most appropriate uses. A one-group pretest–posttest design indicates whether recruitment is feasible, whether the instrument performs adequately, and what the variance of the outcome is, which is the information the later sample size calculation requires.

Regression to the mean cannot be eliminated in this design, so it should be acknowledged. It is most consequential when participants were selected for extreme scores, which is precisely when a follow-up measurement will move back towards the average independently of the intervention.

Yes. Research involving human participants requires ethics committee approval and informed consent, regardless of the simplicity of the design.

References

  1. Campbell DT, Stanley JC. Experimental and quasi-experimental designs for research. Rand McNally, 1963 — the source of the notation and the threats to validity used here. read 2 October 2026. search.worldcat.org/search?q=Campbell+Stanley+experimental+and+quasi-experimental+designs+for+research
  2. STROBE statement, checklist for observational studies. read 2 October 2026. www.strobe-statement.org
  3. EQUATOR Network reporting guideline index. read 2 October 2026. www.equator-network.org

Each source carries the date it was last read. Where a standard moves, this article is revised rather than left up with a note.

How to cite this article

Pick your referencing style and copy it. Your university or target journal may vary the punctuation — check its guide if it has one.

Pubrica Academy. (2026, October 2). Types of Pre-Experimental Research Design. Pubrica. https://pubrica.com/academy/experimental-design/types-of-pre-experimental-research-design/

This article is written to be learned from. Cite it if you are referring to what it says; if you are using the evidence behind it, cite the sources listed above instead.

Pubrica AcademyWritten and maintained by Pubrica’s subject matter experts. Articles are revised when the underlying standards change, and each article states the date on which it was last reviewed.

How Pubrica can help

Where a pilot study is the appropriate next step, a statistical analysis plan specifies what will be measured, and how, before the study begins, and Pubrica’s biostatistics team converts the variance observed in the pilot into the sample size required for the full study.

Find something else

Search the Pubrica Academy

Fifty-eight free articles, templates, guidelines, checklists and worked examples are available without an account. Enter a topic of interest, or open the full library to filter by category, sub-topic or resource type.

Review of a pilot study before the design of the full trial

A scoping assessment is free and takes two business days. It provides a written assessment of what the work requires and which specialist would undertake it.

Get a free quote

Regulations

Check your project against this rule

  1. 1Your work
  2. 2Your details
  3. 3Send
What do you need?

Material you send is held under signed confidentiality, with access limited to the assigned team and encrypted transfer. It is not used to train anything and is not passed to third-party tools. Ask for the agreement before you send anything you would rather cover first.