In many ways, even a brilliant, groundbreaking paper is only as good as its abstract.
That’s because while the average academic paper ultimately has a depressingly small audience (though just how small is a matter of heated debate), the audience for the full paper is even smaller, with most researchers reading only the abstracts of the majority of the papers they cite.
The good news is that a readable, compelling abstract is your best shot at getting a foot in the door with these readers—who include, of course, all those journal editors and reviewers you'd like to impress too. What makes for a great abstract? Here’s a simple approach to crafting one in 2 steps that should take a total of 20 minutes or less.
Step 1:
First, answer each of these questions in no more than 2 sentences:
What (as the background for this paper) is the problem that your entire field is trying to solve?
State it, including the basic details a non-expert would need to understand this paper’s context (e.g., When did the problem start/how long has it been going on? Where is it a problem? Who does it affect? Any core terms that deserve a quick definition? etc.)
How does this paper fit into that problem/struggle to solve it?
How do you tackle the problem differently from past papers? Why was it necessary to do so?
What’s your core thesis or research question?
A good way to start this sentence is “Accordingly,” (that is, in light of the difficulties with past approaches that you described in answering the last question; e.g., “Accordingly, we explore how leaders can achieve x without resorting to those methods that past studies have identified as counterproductive.”)
What theory or approach do you use to test the hypothesis or answer the question?
Mention the theory or approaches that you draw on, build on, extend, adapt, etc. to conceptualize this issue.
What methods did you use to do your actual testing/answering?
If applicable, briefly outline the methods you used to investigate this issue.
What was the main result? Any other results worth mentioning (e.g., unexpected or significant ones)?
State this in terms that any educated adult could understand. In other words, use words that keep their real dictionary meanings where possible (vs. variable labels you developed; e.g., “employees’ trust in their employers” is better than “employee-trust-L” in the abstract).
And finally, what does this result mean for the “real world”?
This is always the best way to end an abstract; yet it’s something that’s often missing from papers entirely. Increasingly, business journals want to see the real-world benefits (typically called “practical implications”) highlighted up front in your paper, if possible.
Step 2:
Next, keeping your answers in the above order, simply connect them with linking words to create excellent flow/readability. Then review your abstract to ensure it meets these finalizing criteria:
Under 200 words
Uses the present tense unless it’s a concrete action that you actually did/completed in the past (e.g., use “We find x” (present tense) because this isn’t one concrete past action, but use “We conducted a survey” because that is a concrete, completed past action).
Explains/defines any key terms for understanding this content, and ideally, use no more than 2–3 terms requiring such explanation
No citations (exception: 1 citation if, and only if, the primary purpose of your paper is to offer a major extension, rebuttal, or counterpoint to the cited article)
The final sentence answers this question: What or who does this paper help, in concrete terms?
Alternatively, if you’re comfortable with generative AI and its many downsides, you can plug your Step 1 answers into your LLM of choice with the following prompt above them:
“Can you please create a 200-word abstract in the style of [target journal] using this information?”
Note: Review the AI-generated abstract VERY CAREFULLY to ensure it captures your intended tone and meaning; I do not recommend this approach and I do not use AI myself—here’s why. However, I recognize that many academics do use it and I think a combination of the right information (like answers to the above questions) and the right prompts can produce decent results—albeit still not nearly as good as the “hand-crafted” ones.
Comentarios