Chapter 7: Research Questions and Hypotheses

Welcome to Last Minute Lecture.

This free chapter overview is designed to help students review and understand key concepts.

These summaries supplement not replaced the original textbook and may not be redistributed or resold.

For complete coverage, always consult the official text.

So,

if I were to ask you, you know, exactly why you decided to listen to this deep dive today, I might actually be ruining my own research.

Yeah, which sounds crazy, right?

But it's true.

In the world of academic research,

the word why carries, well, a ton of dangerous baggage.

It really does.

And by the time we finish this one -on -one tutoring session, you are going to know exactly how the very words you choose to ask a question completely shape the reality you are able to discover.

Exactly.

So, grab a coffee, get comfortable, and just take a deep breath.

Yeah.

If you're staring down a massive research design assignment right now, just consider us your late -night study buddies.

We've done the heavy lifting, we've color -coded the notes, and we're going to help you master this material from chapter seven.

We definitely are.

And the core concept we're tackling today is how researchers guide their readers through a study using what methodologists call signposts.

Right, signposts.

Yeah.

Think of your research project like a dense overgrown forest.

The very first signpost you plant in the ground is your broad purpose statement.

But you obviously can't just stop there.

No, of course not.

You'd get lost.

Right.

So, you have to carve a specific path through those trees.

And you do that by narrowing your broad purpose down into the true stars of today's session, which are research questions and hypotheses.

I love that visual, honestly.

Carving a path.

And to make sure you have a perfect map for your assignment, we are going to structure this session really logically.

Definitely.

It helps to follow the exact roadmap from the text.

Yeah.

So, we will start with the highly structured rigid rules of quantitative research.

Then we'll take a sharp turn into the open exploratory world of qualitative research.

And finally, we will look at how to masterfully combine them in the cutting edge realm of mixed methods research.

Okay.

Let's unpack this.

We're starting with quantitative questions and hypotheses.

Now, if the goal is to narrow down a broad purpose, quantitative research is the most rigorous structured way to do exactly that.

Right?

Absolutely.

It relies on strict variables and really precise predictions.

Right.

So, to put it in perspective, is it fair to say a quantitative hypothesis is kind of like a courtroom trial?

You know, that is actually a brilliant way to frame it.

Think about how a trial works.

Before you ever start examining the evidence or calling witnesses,

the prosecution has to state their exact prediction.

Like they have to read the specific chart.

Exactly.

You cannot just walk into a courtroom and say,

let's look at all the evidence and see what crime might have organically happened.

You have to declare, I predict the defendant committed this exact crime in this exact way.

Right.

You have to call your shot before you swing the bat.

You are locking in a prediction.

You are.

And in quantitative studies, those predictions are specifically about the relationships between variables.

Which is huge.

If you are prepping for an exam, you absolutely need to know how to identify these variables.

100%.

So, you typically have an independent variable, which is the condition you are manipulating, or like the way you're grouping your participants.

And then you are looking to see the impact of that independent variable on a dependent variable, which is your outcome.

Yeah.

And just to add a slight nuance for your notes, if you are doing a survey or an observational study where you aren't actively manipulating anything,

researchers usually call them predictor variables and outcome variables.

Okay.

Predictor and outcome.

That makes sense.

And they are also mediating variables.

These are fascinating because they sit right between the predictor and the outcome.

They explain the actual mechanism of how the effect happens.

Wait, can you give an example of that?

Sure.

So, if your predictor is hours studying and your outcome is exam score,

your mediating variable might be retention of information.

The studying causes the retention, which in turn causes the high score.

Okay.

Got it.

But a researcher doesn't just pull these variables out of thin air to see what happens, right?

No, definitely not.

Yeah.

The consensus in the field is that the most rigorous quantitative research is grounded in existing theory.

Like you take a published, established theory from the literature and your hypotheses either logically flow from it to test it, or they explicitly challenge it.

Precisely.

You are testing theory.

And when it comes to the actual predictions, the hypotheses, there are two fundamental forms you really need to recognize.

Okay.

What's the first one?

The first is the null hypothesis.

Null essentially means zero.

Right.

Nothing.

Exactly.

A null hypothesis makes a formal prediction that in the general population, there is absolutely no relationship or no significant difference between the groups you are studying.

So it usually reads something like, there is no difference in exam scores between group A and group B.

Yes.

And the entire statistical intent here is actually to try and disprove or reject this expectation.

Wow.

So it's almost like assuming innocence until proven guilty.

That's a great way to put it.

But most of the time, researchers are actually focused on the second type, the alternative hypothesis.

And more specifically, the directional alternative hypothesis, right?

Yes, exactly.

This is where the researcher confidently predicts that there is a significant relationship, and they predict the exact direction it's going to go.

So they use highly specific directional verbs, words like affect, influence, predict, impact, determine, or cause.

Right.

So instead of a weak statement like diet and weight are related, a directional alternative hypothesis plants a flag.

It takes a stance.

Yeah.

It says there's a positive relationship between caloric intake and weight, such that individuals who consume more calories are more likely to have a higher body mass.

You are explicitly predicting the direction of the relationship based on your underlying theory.

Now, a lot of students get hung up on exactly how to write these out.

But methodologists actually use very standard formulas for this, don't they?

They do.

It's not about being a creative writer at all.

It's about being clear.

Right.

If a researcher just wants to describe a single variable, they use a descriptive script.

They basically ask for the frequency and variation of scores on a specific variable for a specific group of participants, like in example 7 .1 in the text.

Exactly.

The data analysis for a descriptive question is purely foundational.

You are looking at means, standard deviations, and the range of scores.

But most quantitative research wants to relate variables together, right?

Yeah.

For that, you use a relationship -oriented formula like example 7 .2.

You start by explicitly naming the theory you are using.

Then you explain what that theory posits about the relationships between your variables.

And then you state the prediction.

Right.

You say something like, it is therefore predicted that there will be a relationship between this predictor and this outcome, such that, and then you explain the exact direction of the effect.

Let's ground this in a real piece of published research.

There is that classic study in the text, example 7 .3, by a researcher named Moore from 2000.

Oh yes, the study on social identities in Israel.

Exactly.

She investigated the social identities of religious and molecular Jewish and Arab women in Israel.

Moore didn't just ask, how did these women feel about politics?

No, she used highly specific directional hypotheses.

Right.

One prediction was that religious women with a strong gender identity would be less sociopolitically active than secular women.

And another predicted that the relationships between gender identity, religion, and social action would be weaker among Arab women compared to Jewish women.

Yeah.

Listen to the very specific words chosen there, less active, weaker relationships.

Those are the rigid directional signposts we are talking about.

And it is crucial to understand why quantitative researchers are so obsessed with these strict directional predictions.

Why go through all this trouble?

Because in quantitative research,

you are almost always analyzing a relatively small sample of people.

True.

But by testing these highly formalized hypotheses using rigorous statistical procedures, the investigator earns the right to draw mathematical inferences about an entire massive population.

That's wild.

You can study 500 people and make a statistically valid claim about 5 million people.

Exactly.

That is the incredible power of the quantitative approach.

Okay.

Well, contrast is the absolute best engine for learning.

So if quantitative research is all about locking in a strict, unyielding prediction before you even collect a single piece of beta, what happens when a researcher absolutely does not want to predict the outcome?

Which happens a lot.

Right.

What if predicting the outcome would actually ruin the study because they just want to explore what is happening organically?

Well, that organically leads us to the second part of our roadmap, which is qualitative research.

Yeah.

In qualitative design, you throw hypotheses out the window completely.

You predict nothing.

Nothing.

And that is a major paradigm shift for a lot of students.

Instead of testing objectives or predicting the movement of variables,

qualitative inquirers simply state open -ended research questions.

Oh, wait.

I have to push back on a major rule in qualitative methodology here.

Okay.

What's that?

The standard framework insists that qualitative questions should begin with what or how.

But Strongly warn against ever using the word why.

Feels counterintuitive to me.

I get that.

Isn't the entire point of academic research to ask why things happen?

Why is why suddenly a bad word?

What's fascinating here is that the word why carries intense hidden baggage that most of us just don't realize.

Baggage?

Like what?

Well, when you ask why something occurs, it strongly implies that you are hunting for probable cause and effect relationship.

And as we just established, cause and effect thinking is the domain of quantitative variable.

Exactly.

If you sit down with a participant and ask them why they did something, you are inadvertently forcing them to rationalize their behavior through the lens of your preconceived variables.

But if you ask what happened or how they experienced it, you convey an open emerging design.

You hand the microphone over to the participant, allowing their unfiltered lived reality to emerge without boxing them into your academic assumptions.

That makes total sense.

You are interrogating them.

You are inviting them to share.

Precisely.

So how do we structure these what and how invitations?

Because the field relies on a very specific hierarchy, right?

You don't just write a list of 20 random questions.

No, definitely not.

You start with one or two central questions.

This is the absolute broadest, most overarching question you can possibly ask about the central phenomenon you are exploring.

Okay.

So that's the big umbrella.

Yes.

And once you have that massive umbrella, you follow it up with a small handful, usually five to seven specific sub -questions.

Which narrows the focus into digestible chunks, but they still leave the exploration wide open.

Right.

And researchers like Miles and Huberman have famously recommended that qualitative researchers should write no more than a dozen research questions in total, combining the central and sub -questions.

Because if you have more than 12, you've probably lost your focus.

Exactly.

And those sub -questions are incredibly practical, by the way.

They eventually form the backbone of your actual data collection.

They become the core prompts you use during an interview protocol or the observational checklists you use in the field.

Got it.

And just like quantitative research relies on directional words like predict and cause,

qualitative research relies heavily on exploratory verbs.

Yes.

Very different vocabulary.

If you are writing a qualitative question, you want to use verbs like discover, describe, understand, or explore.

And the way you phrase your question actually changes depending on the specific qualitative tradition you are using.

Right.

Because qualitative research isn't a monolith.

Exactly.

That is a crucial distinction for anyone designing a study.

So if you are conducting an ethnography, which is the deep study of a culture -sharing group, your questions might involve asking how a specific community uses native language to maintain social hierarchies.

Okay.

What about phenomenology?

Phenomenology seeks to understand the universal essence of a lived experience.

So you might ask what participants experienced and in what specific contexts.

Think of a heavy, profound question like, what is it like for a mother to live with a teenage child who is dying of cancer?

Oh, wow.

Yeah, you weren't looking for variables there.

You were looking for the phenomenological essence of grief and care.

Exactly.

And if you're doing grounded theory, your questions are totally different again.

You are directing your inquiry toward generating a brand new theory about a process.

So you might ask how caregivers and patients interact over time to negotiate a treatment plan.

Right.

And lastly, if you're doing a case study, you are looking for an in -depth description of a very specific bounded case, looking for the unique themes that emerge from it.

But regardless of whether you are doing an ethnography or a case study, there is one golden kind of terrifying rule for qualitative questions.

Oh, yeah.

You must expect them to evolve and change while you were conducting the study.

Which absolutely drives quantitative researchers crazy.

It really does.

In a quantitative study, changing your hypothesis halfway through data collection is a massive ethical violation.

It's called p -hacking.

Right.

But in qualitative research, continually reviewing and reformulating your central question based on what your participants are teaching you is actually the hallmark of a rigorous responsive study.

It shows you're actually listening.

So let's look at how the formula for a central question actually works.

You don't need a complex paragraph.

The standard approach simply asks how or what is the central phenomenon for these specific participants at this specific research site.

It is beautifully simple.

It just sets the stage.

Let's walk through some fascinating real world studies to see this in action.

Take example 7 .4, the ethnography by Mack and Gayle and Haywood.

Right.

They spent time with working class Pakistani and Bangladeshi young men in Britain.

Exactly.

Their central question didn't make any assumptions about race or class outcomes.

It simply asked what these young men's core beliefs were regarding ethnicity and religion and how they constructed their everyday experiences of family and social life in a rapidly changing Britain.

It perfectly zeroed in on the phenomenon, which was their core beliefs, the participants, the young men, and the site,

Britain.

Or imagine you want to understand how elderly residents interact with young children.

There is a wonderful case study, example 7 .5 by Hernandez, and a team of researchers who looked at an intergenerational play group set up inside a residential aged care facility.

Oh, I love this study.

Me too.

Their central question was elegantly broad.

It was just how do participants engage in an intergenerational play group within the context of a residential aged care facility?

The phenomenon they were studying was simply engagement.

Right.

Because they left it that broad, they were able to organically discover deeply moving, unexpected themes from the participants, like the concept of learning from each other, or the profound joy of appreciating experience in the moment.

Which is incredible.

If they had gone in with a rigid quantitative hypothesis, they probably would have just counted how many times a child handed a toy to a senior citizen and entirely missed the magic of what was actually happening.

That is the perfect illustration of qualitative power.

It really is.

And to understand how sub -questions work, we could look at example 7 .6, the study by Creswell and Bays.

They wanted to tackle a massive, complicated topic.

What is the campus climate toward diversity?

I mean, that central question is almost too big to answer.

Exactly.

So they masterfully broke it down with targeted sub -questions.

They asked, what are the students' attitudes?

How is diversity actively encouraged by the administration?

How is it integrated into classroom instruction?

How is it handled by the campus police?

So they subdivided a giant nebulous phenomenon into highly specific, manageable interview topics that all rolled up to answer the central question.

Exactly.

So we've covered the two extremes.

You now understand how to define variables to predict outcomes in quantitative research, and how to use exploratory verbs to uncover open -ended experiences in qualitative research.

But what if your specific research problem is so complex that you desperately need both the numbers and the stories?

Here's where it gets really interesting.

How do we execute both approaches in a single study without creating an absolute chaotic mess?

Welcome to the frontier of mixed methods research.

Yeah.

To use an analogy, mixed methods is not like a TV dinner.

In a TV dinner, you have the quantitative peas in one compartment and the qualitative carrots in another compartment, and they never touch.

Right.

They're separate.

But mixed methods is a stew.

You are throwing everything into the pot, and the mixed methods question is designed to evaluate the brand new flavor that is created by integrating those ingredients together.

I love the stew analogy because it highlights the transformation of the data.

The mixed methods question is a relatively new, innovative concept in research methodology.

It is.

When you sit down to write one, you are doing something radical.

You aren't just asking about the subject matter of your study anymore.

You are explicitly asking a methods question.

Wait, asking it about the methods?

Yeah.

You are asking what new insight is learned by specifically integrating or combining the rigid quantitative numbers with the rich qualitative stories.

Methodologists emphasize that if you are claiming to do a mixed methods study, you must advance all three types of questions in your paper, right?

Absolutely.

You need your quantitative hypotheses to rigorously analyze the statistical numbers.

You need your qualitative questions to deeply analyze the human stories.

And then crowning the entire design, you need this special third question, the mixed methods question, to explicitly address the integration of the two databases.

And the specific order in which you present these three sets of questions depends entirely on the architectural design of your study.

Let's break this down clearly.

Okay, let's do it.

If you are using a convergent design, which means you are collecting both the numbers and the stories at the exact same time and merging them, you can present either the quantitative or qualitative questions first.

It doesn't really matter.

Because they're happening concurrently.

Right.

But if you are using an explanatory sequential design, well, the name kind of gives it away.

You collect the quantitative data first to find the statistical trends, and then you use qualitative interviews to explain those numbers.

Ah, so in that case, your quantitative hypotheses must be presented first.

Exactly.

Conversely, if you are using an exploratory sequential design, you flip the script.

You explore qualitatively first because you don't even know what variables exist yet.

So you talk to people to build a framework, and then you test that framework quantitatively.

Yes.

Naturally, your qualitative questions take the lead there.

That makes total sense.

Now, when you write the actual mixed methods question, the field requires you to clearly state two things, your intent and your procedure.

Right.

Are you trying to compare the data or use one to build upon the other?

And are you merging the data together or embedding one inside the other?

Let's look at example 7 .7, a brilliant application of this by a researcher named Mosholm and her colleagues.

Oh, the health sciences study.

Yeah.

They focused on patients who are presenting at a hospital with non -specific, vague symptoms that doctors suspected might be cancer.

It is obviously a terrifying, uncertain time for a patient.

Absolutely.

The researchers wanted to deeply understand the patient's health -related quality of life during this diagnostic waiting period.

So they deployed a quantitative statement.

Their goal was to statistically measure changes in quality of life using standardized clinical metrics.

And then they deployed a qualitative statement.

They wanted to deeply describe the patient's emotional and physical experiences through interviews.

So you have the hard numbers and you have the tears in the waiting room.

Exactly.

And then they brought it all together with a masterful mixed method statement.

They explicitly stated their intent was to merge these two findings to obtain a far more comprehensive understanding of quality of life than either the clinical metrics or the interviews could provide alone.

By explicitly stating they were merging the data for a comprehensive understanding, they perfectly executed a convergent design.

They really did.

Now, to see why this is so difficult to get right, let's look at a different study, Example 7 .8, that actually drew a gentle critique from methodologists.

Great.

The study by D 'Cuito and Esteta.

They conducted a fascinating study on how STEM teachers adapted to online teaching during the pandemic.

Yeah.

They used quantitative surveys to figure out what digital tools the teachers were using.

But they also conducted qualitative interviews to hear the teachers reflect on how their assessment models were changing.

It was undeniably a rigorous, highly valuable study.

The results were excellent.

But methodologists point out a structural flaw.

The authors never clearly labeled their qualitative questions, nor did they write an explicit mixed methods question.

Right.

In their conclusion, they successfully combined the hard data about assessment modes with the rich interview data about teacher reflections.

But the critique is that they should have provided a clearly labeled upfront mixed methods question that highlighted their specific intent to combine that reflective qualitative data with the quantitative assessments.

Exactly.

I can hear a stressed out college student right now asking

if the study was rigorous and the results were great, who cares if they forgot to label a specific question at the beginning?

It's a fair point.

Yeah.

But if we connect this to the bigger picture, why does the academic community insist on such rigid formatting for the mixed methods question?

Because executing a true mixed methods design is absolutely exhausting.

It really is.

You're essentially doing two entirely separate full -scale studies and then undertaking a massive third analytical step to synthesize them.

So writing a distinct mixed methods question forces the researcher to pause and justify why they are going through the exhaustive, expensive effort of collecting two entirely different types of databases.

Right.

If you cannot clearly articulate what unique insight you are going to gain by integrating the numbers and the stories, you probably shouldn't be doing a mixed methods study in the first place.

Precisely.

It proves to the peer reviewers, to your academic advisors, and most importantly to yourself, that the final stew is actually greater than the sum of its individual ingredients.

Well said.

So what does this all mean?

Let's take a breath and quickly recap the journey we've taken in this tutoring session.

Sounds good.

We started deep in the rigid, structured signposts of quantitative research where you must define your variables, root yourself in theory, and lock in your directional hypotheses to make statistical predictions about large populations.

Then we dropped the predictions entirely and wandered into the open -ended what and how of qualitative central and subquestions, utilizing exploratory verbs to deeply understand the nuance of human experiences.

And finally, we brought those two worlds together into the integrated stew of mixed methods, where the mixed methods question explicitly demands to know what new profound understanding is gained by merging the numbers and the stories.

If you've absorbed this, you now have the exact frameworks and philosophical understanding you need to tackle your assignment.

You understand how the foundational assumptions of a researcher shape the questions they ask, how those questions dictate the architectural design choices, and how those choices ultimately govern the data collection and the final analysis.

Absolutely.

But before we let you get back to your writing, we want to leave you with a philosophical thought that naturally arises from mastering this material.

Okay, late on them.

If the very words we choose to ask a question, you know, choosing predict versus explore or insisting on how instead of why, actually dictate the methodologies we are allowed to use,

do our research questions actually shape the reality we are able to discover?

That is a deep question.

Right.

Think about that as you design your own studies.

If we only ever look for rigid variables and strict cause and effect, do we completely miss the messy, beautiful nuance of the human experience?

And conversely, if we only ever collect open -ended individual stories, do we miss the overarching hidden mathematical patterns that govern entitled populations?

Exactly.

The questions you choose to write on your paper tonight don't just guide your study.

They define the absolute limits of what you are capable of learning.

It is a profound responsibility for any researcher.

The signposts you choose to place in the forest ultimately determine the destination you can reach.

Well, on behalf of the last -minute lecture team, we explicitly want to thank you for spending this time with us.

We know research design can feel incredibly heavy and overwhelming, but you have the tools now.

You really do.

Best of luck applying these concepts to your own research designs and acing that big assignment.

You've got this.

You are going to do great.