How to Write Research Objectives and a Hypothesis: The Blueprint of Your Study

Three Weekends and One Objective

Dr. Bushra Fatima placed her CPSP synopsis draft on the table with the confidence of a woman who had spent three weekends writing it.

Dr. Muhammad Yaqoob picked it up and turned to the objectives section. He read it once. Then he read it again. Then he set it down very gently, which everyone in the Research Room at UPMED Hospitals had come to recognise as a sign that something needed significant improvement.

Dr. Yaqoob: “Bushra, what is your general objective?”

Dr. Bushra: “To study the effects of antenatal steroids on neonatal outcomes in preterm deliveries at UPMED Hospitals.”

Dr. Yaqoob: “And your specific objectives?”

Dr. Bushra: “I only have one. It says: to find the outcomes.”

Silence.

Dr. Hammad Ali: “That is… very specific. Very vague. Both at the same time.”

Dr. Bushra: “I thought it was fine!”

Dr. Yaqoob: “It is not terrible. You have identified your topic. But objectives and hypotheses are where synopses are won or lost. A CPSP reviewer reads your objectives and decides in thirty seconds whether your study is worth approving. Let us fix this together, and properly.”

He picked up a marker and wrote three words at the top of the whiteboard: OBJECTIVE. SPECIFIC OBJECTIVE. HYPOTHESIS.

“These are three different things,” he said. “Most doctors treat them as one. That is the first mistake.”

What Is a Research Objective, and Why Does It Matter?

A research objective is a clear, concise statement of what your study intends to achieve. It tells the reader, whether a CPSP synopsis reviewer, a university thesis examiner or a journal editor, exactly what you are setting out to do.

In Pakistani medical research, objectives are usually written at two levels:

• The general objective (also called the aim): one broad statement describing the overall purpose of the study.
• The specific objectives: two to five focused, measurable sub-aims that together fulfil the general objective.

The general objective sets the direction. The specific objectives describe the individual steps you will take to get there.

The General Objective: One Sentence, One Direction

Your general objective is a single sentence. It usually begins with an infinitive verb (to determine, to assess, to evaluate, to compare), followed by what you are studying, in whom, and where. The formula is simple:

To [verb] the [outcome / relationship / frequency] of [exposure / condition] in [population] at [setting / time period].

Weak: To study antenatal steroids in preterm deliveries.

Strong: To compare neonatal respiratory outcomes between preterm neonates (less than 34 weeks of gestation) born to mothers who received a complete course of antenatal corticosteroids and those whose mothers did not, at UPMED Hospitals, Lahore, over three months after the approval of the synopsis.

The strong version tells you what (a complete course of antenatal corticosteroids versus none), which outcome (neonatal respiratory outcomes), who (preterm neonates under 34 weeks), where (UPMED Hospitals) and when (three months after the synopsis is approved). A reviewer can picture the study at once. Notice also the verb. Bushra is not giving the steroids herself; she is observing what happened. So “to compare” fits her study better than “to determine the effect”, which sounds like a trial.

Dr. Sumaira Talib: “Sir, does the general objective have to mention the study design?”

Dr. Yaqoob: “Not necessarily, though some synopsis formats ask for it. Focus on what you are measuring. The design lives in its own section of the synopsis.”

Specific Objectives: Breaking the Aim into Measurable Steps

Each specific objective should be:

• Specific: focused on one variable or relationship.
• Measurable: something that can be counted, compared or tested statistically.
• Achievable: realistic within your design and sample size.
• Relevant: directly linked to your general objective.
• Time-bound: completed within your study period.

This is the SMART framework applied to research objectives. For a CPSP synopsis, two to four specific objectives are ideal. More than five, and reviewers wonder whether your study is trying to do too much. Too few, and they wonder whether you have thought it through.

The Verb Matters

The verb in each specific objective tells the reviewer which analysis you intend, even before they reach your methodology:

Verb in the objective	What it implies	Usual analysis
To determine the frequency of	Description	Frequency, percentage, mean, SD
To compare [X] between two groups	Group comparison	Chi-square, t-test, Mann-Whitney U
To find the association between [X] and [Y]	Association	Chi-square, odds ratio, relative risk, correlation
To identify the factors associated with [outcome]	Several exposures together	Binary logistic regression
To determine the diagnostic accuracy of [test]	Diagnostic study	Sensitivity, specificity, predictive values

Dr. Hassan Raza: “Sir, so if I write ‘to study the outcome after ORIF’, that is a bad objective?”

Dr. Yaqoob: “It tells us nothing about how you will measure it. Better: ‘To determine the mean Patient-Rated Wrist Evaluation score at three months after ORIF of distal radius fractures in adults at UPMED Hospitals, Lahore.’ Now I know your patients, your measurement tool and your time point. You have one group and no comparison, so you need no hypothesis.”

Dr. Hassan: “That is very long.”

Dr. Yaqoob: “Precision is not the enemy of brevity. In research writing, vagueness is the enemy.”

Bushra’s Revised Objectives

After twenty minutes at the whiteboard, Bushra’s specific objectives read:

• To compare the frequency of neonatal respiratory distress syndrome (RDS) between preterm neonates born to mothers who received a complete course of antenatal corticosteroids and those whose mothers did not.
• To compare the need for mechanical ventilation within 24 hours of birth between the two groups.
• To compare the duration of NICU stay (days) between the two groups.

Dr. Yaqoob had also removed her fourth objective, on maternal factors linked to completing the steroid course. “That is a different research question,” he said. “Keep it for another study.”

Bushra read the three objectives back. “They are much more specific now. But also much harder to achieve.”

Dr. Yaqoob: “That is the honest trade-off of good research. Vague objectives are easy to write and impossible to publish. Specific objectives are harder to write and far more likely to be approved and published.”

The Hypothesis: Your Testable Prediction

A hypothesis is a specific, testable prediction about the relationship between variables, stated before you collect any data. It comes in two forms:

• The null hypothesis (H₀): there is no difference or relationship in the population. This is what your statistical test tries to reject.
• The alternative hypothesis (H₁): there is a difference or relationship. This is what you expect to find.

If your data give a p-value below the significance level (usually 0.05), you reject H₀ and the data support H₁. If not, the result is “not statistically significant”. That is not a failure, and a well-conducted null result is still publishable. But word it carefully: “no evidence of a difference”, never “no difference”. A small study may simply lack the power to detect a real effect, so report the confidence interval too. We will come back to p-values and confidence intervals in the statistics part of this series.

Dr. Junaid Rashid: “I never understood when a CPSP synopsis needs a hypothesis. Some have one, some do not.”

Dr. Yaqoob: “There is a clear rule. A descriptive study that only measures a frequency in one group needs no hypothesis. A comparative study states its alternative hypothesis, in a non-directional form. Sumaira only measures the frequency of surgical site infection, so she needs none. Hammad compares two pneumoperitoneum pressures in a trial, so he writes one. Bushra compares two groups of neonates, so she writes one for each comparison.”

How to Write a Good Hypothesis

A good hypothesis names the outcome, the groups being compared and the population:

There is a difference in [outcome] between [exposed group] and [unexposed group] in [population].

Weak: Antenatal steroids have an effect.

Strong: There is a difference in the frequency of neonatal respiratory distress syndrome between preterm neonates born to mothers who received a complete course of antenatal corticosteroids and those whose mothers did not.

The strong version contains no p-value. A hypothesis is a statement about the population; the significance level and the statistical test belong in the statistical analysis section. It is also non-directional: it says the frequencies differ, not which group will be lower.

Directional or Non-Directional?

A directional (one-tailed) hypothesis predicts which way the effect goes: “Group A will have less RDS than Group B.” A non-directional (two-tailed) hypothesis only predicts that a difference exists.

For CPSP and university synopses, use a non-directional hypothesis and a two-tailed test. One-tailed tests are rarely justified in medical research, and reviewers distrust them because they make a significant result easier to reach.

Dr. Zunaira Malik: “Sir, I wrote that areca nut chewing will be more common in patients with oral submucous fibrosis than in controls. Is that directional?”

Dr. Yaqoob: “Yes. Write it as non-directional and use a two-tailed test. If chewing turned out to be less common in your cases, a one-tailed test could not detect it at all. Strong literature on one side is not enough reason to ignore the other side.”

Common Mistakes, With Fixes

• Objectives that cannot be measured. Wrong: To understand the impact of diabetes on patients. Right: To compare mean HbA1c at six months between patients on metformin alone and those on combination therapy.
• Too many specific objectives. Eight objectives usually means several research questions in one synopsis. Choose your main question and stay with it.
• A hypothesis that does not match an objective. If the objective is “to compare X between groups A and B”, the hypothesis is “there is a difference in X between groups A and B”.
• P-values inside the hypothesis. State the significance level in the statistical analysis section, not in the hypothesis.
• Objectives in the past tense. A synopsis is written before the study: “to determine”, not “to have determined”.

At a Glance

	General objective	Specific objectives	Hypothesis
Purpose	Direction of the study	Measurable sub-aims	Testable prediction
How many	One	Two to five	One per key comparison
Contains statistics?	No	No	Groups and outcome, no p-value
When needed	Always	Always	Comparative studies only, non-directional

The Synopsis That Got Approved

Three weeks later, Bushra came back to the Research Room with a printed email from the CPSP Research Evaluation Unit. Her synopsis had been approved, with minor comments, mostly about the sample size calculation. (“We will cover sample size in its own session,” said Dr. Yaqoob, making a note.)

Dr. Bushra: “They specifically wrote that the objectives were well-defined and measurable. Nobody has ever said that about my work before.”

Dr. Yaqoob: “Because your objectives were never well-defined and measurable before.”

Dr. Bushra: “Sir, that is honest.”

Dr. Yaqoob: “I try to be. Research requires honesty: with your data, with your reviewers, and first of all with yourself.”

Hammad was already rewriting his own objectives. Hassan was copying the verb table into his phone. Junaid had not said a word for twenty minutes. He was writing.

Key Takeaways

• The general objective is one sentence of purpose. Specific objectives are two to five measurable sub-aims. A hypothesis is a testable prediction about the population.
• Use the formula: to [verb] the [outcome] of [exposure] in [population] at [setting and time].
• The verb signals the analysis: “to determine the frequency” means description, “to compare” means a group comparison, “to find the association” means odds ratios or relative risks.
• Descriptive studies need no hypothesis. Comparative studies state a non-directional alternative hypothesis and use two-tailed tests.
• Keep p-values out of the hypothesis, and if you cannot reject H₀, write “no evidence of a difference”, never “no difference”.