01 · The Question
Are Some Research Objective Verbs Actually Wrong?
Students are often told never to write objectives such as:
To know students' perceptions of generative AI.
To understand how students use generative AI.
Then they are advised to replace them with:
To determine students' perceptions of generative AI.
The replacement certainly sounds more research-like. But has the objective actually improved?
Sometimes yes. Sometimes not. Know and understand can leave the intended research accomplishment unclear, particularly in specific objectives. Yet determine can be equally vague when it is attached to an undefined phenomenon. The quality of an objective depends on the meaning of the complete statement, not on whether its first verb has been granted methodological respectability.
03 · What You Need to Know
The Problem Is Usually Vagueness, Not a Forbidden Word
Why “To Know” Is Usually a Poor Choice
Consider:
To know the factors affecting students' use of generative AI.
What does “know” require the researcher to do? Identify factors reported by participants? Estimate statistical associations? Develop a causal explanation? Compare predefined predictors?
The verb does little to specify the intellectual task.
Research-methods guidance commonly recommends action verbs that make objectives specific enough to evaluate. WHO-derived training materials, for example, recommend verbs such as determine, compare, verify, calculate, describe, and explain, while cautioning against vague formulations such as appreciate, understand, believe, and study.
“To know” presents a similar problem: knowing is the researcher's cognitive state rather than a clear description of the research accomplishment.
“To Understand” Is More Complicated
Understand is frequently criticized because it can be difficult to evaluate without further specification.
For example:
To understand generative AI use among university students.
This says very little about what evidence will be collected or what aspect of use matters.
But consider:
To understand how first-generation university students experience the use of generative AI while developing their identities as academic writers.
The statement is substantially more informative. In some qualitative and interpretive traditions, understand may also reflect the epistemic purpose of the inquiry more faithfully than a verb such as measure.
The issue is therefore not simply whether the word appears. The question is whether the objective identifies a sufficiently clear phenomenon, population, context, and form of inquiry.
Why “To Determine” Is Popular
Determine appears frequently in lists of research-objective action verbs and can be entirely appropriate.
For example:
To determine the prevalence of weekly generative AI use for academic writing among undergraduate students.
The intended research accomplishment is reasonably clear. The study seeks to estimate a defined quantity in a defined population.
However:
To determine the impact of artificial intelligence on education.
remains problematic. Impact is undefined, the relevant form of AI is unclear, the educational context is enormous, and the intended inferential claim is unspecified.
The verb has not fixed any of those problems.
“Determine” Can Also Overstate Certainty
Researchers should be particularly careful when determine is paired with causal or definitive language.
Suppose a cross-sectional survey examines self-reported generative AI use and writing self-efficacy.
More defensible: To examine the association between frequency of generative AI use and academic writing self-efficacy.
Potentially misleading: To determine the effect of generative AI use on academic writing self-efficacy.
The second statement does not become methodologically stronger because it uses determine. It may instead promise an effect that the design cannot establish.
Watch Out
Do not replace a vague verb with a more decisive one if the replacement changes the type of claim the study makes. Stronger wording is not stronger methodology.
Choose a Verb That Names the Actual Research Task
| Vague or Generic Version |
Ask What You Really Mean |
Possible Revision |
| To know students' AI use |
Do you want to document patterns of use? |
To describe students' patterns of generative AI use in academic writing. |
| To understand students' perceptions |
Do you want to explore how they interpret a particular experience? |
To explore students' perceptions of how generative AI influences academic writing. |
| To determine differences |
Are you comparing defined groups? |
To compare verification practices between students with and without formal AI training. |
| To determine the relationship |
Are you examining an association? |
To examine the association between AI literacy and verification practices. |
| To know how common AI use is |
Are you estimating prevalence? |
To estimate the prevalence of generative AI use for academic writing. |
| To understand student experiences |
Is the inquiry intentionally interpretive? |
To explore how students experience generative AI use during academic writing. |
The point is not that the right-hand verbs are universally superior. They are more informative because each one names the research task more directly.
Do Not Confuse Research Objectives With Learning Objectives
The prohibition against know and understand is especially familiar from instructional design. In learning objectives, these verbs are often discouraged because educators need observable evidence that a learner has achieved the intended outcome.
Research objectives are different. They describe what a study intends to accomplish.
Some of the same concerns about vagueness apply, but the rules should not simply be imported wholesale. A qualitative study may legitimately seek understanding as an epistemic goal, even if its specific objective is better expressed with language such as explore, interpret, or examine.
This distinction matters when deciding whether research objectives require measurable action verbs. Assessability matters, but it does not require every research objective to resemble a classroom learning outcome.
Context Can Rescue a Broad Verb, but Precision Is Usually Better
Suppose the research question, methodology, and surrounding text make the intended meaning perfectly clear. In that context, an objective using understand may be interpretable.
Still, objectives are often read independently in abstracts, protocols, proposal summaries, ethics applications, tables, or evaluation documents. Clear wording therefore has practical value.
If understand can be replaced by a more precise verb without distorting the methodology, the revision usually improves traceability.
“To Determine” Works Best When What Is Being Determined Is Explicit
Consider these examples:
To determine the prevalence of...
To determine whether the prespecified eligibility criteria are met...
To determine the diagnostic accuracy of...
To determine the optimal threshold under the stated criterion...
In each case, the object gives determine a reasonably precise meaning.
By contrast:
To determine student performance.
To determine the effect of technology.
To determine the challenges of AI.
remain underspecified because the substantive concepts and analytical relationships are unclear.
“Understand” May Be More Defensible in Exploratory Research
An exploratory study does not necessarily begin with variables whose relationships have already been operationalized. It may seek to develop an account of experiences, practices, meanings, mechanisms, or processes.
For that reason, insisting that every exploratory objective use verbs associated with quantitative measurement can be counterproductive.
Still, “to understand” by itself rarely tells readers enough. A clearer formulation may specify what is being understood and through what kind of intellectual task:
Broad: To understand faculty experiences with AI policy.
Clearer: To explore how faculty members interpret and respond to institutional generative AI policies in assessment.
The revised objective remains exploratory while making the inquiry more visible.
The Whole Objective Must Be Achievable
Whether you write know, understand, determine, explore, or evaluate, the final question remains methodological:
Can the study actually generate evidence capable of addressing this statement?
A precise verb attached to an impossible objective does not help. The objective still needs to be achievable with the study design.