Chapter 3: The Use of Theory

Welcome to Last Minute Lecture.

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

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

For complete coverage, always consult the official text.

You know, usually when you picture a theory, you just imagine this abstract concept floating up in the clouds somewhere.

Oh, absolutely.

I mean, we tend to view it as this highly intimidating, impenetrable wall of text.

Right, like dusty academics and tweed jackets arguing over philosophy that has literally zero connection to the real world.

Yeah, just something to make a syllabus heavier, really.

Exactly.

But then you actually step into the world of research design and suddenly theory isn't in the clouds at all.

No, it really isn't.

It's like the heavy steel cable holding a suspension bridge together.

I mean, without it, the whole structure just completely collapses into the river.

That is a great way to put it.

So if you are a college student staring down a massive research method syllabus right now, you are in the exact right place.

This deep dive is entirely for you.

We are gonna rip apart chapter three, the use of theory from your textbook research design.

Yep, and we're gonna show you exactly why your research project will fail without a solid theoretical foundation.

Because we wanna walk you through the entire lifecycle of a research project today.

Right.

We'll look at how your foundational assumptions dictate the specific questions you asked.

And then how those questions shape your design choices.

Exactly, which ultimately dictates the data you collect and how you interpret it.

So to do that, we're gonna start by defining theory in its most rigid mathematical setting, which is quantitative research.

And then we'll explore what I like to call the inductive flip, how qualitative researchers use theory in completely the opposite way.

Oh, I like that.

And finally, we'll look at how to blend those numbers and narratives together in mixed methods research.

So to really grasp how theory is used, we have to establish a working definition first.

Right, we need a baseline.

The textbook pulls a classic definition from Kerlinger.

This was written back in 1979.

Okay.

He defined a theory as a set of interrelated constructs or variables that presents a systematic view of phenomena.

So it specifies the relations among those variables, right?

Yes, with the exact purpose of explaining and predicting natural phenomena.

Okay, the textbook uses a visual to explain this.

I really wanna unpack for a second.

They call it the rainbow metaphor.

Ah, yes, the rainbow.

Yeah, so imagine an independent variable planted firmly in the ground on one side and a dependent variable planted on the other.

Okay.

The theory is the bright rainbow that arches across the sky, bridging the two together.

Like it's the overarching explanation for how and why those variables connect.

It's a helpful starting point for sure.

But let's make that rainbow a bit more mechanical.

Make it mechanical.

Yeah, because theory does actual heavy lifting.

I mean, a rainbow is pretty, but it's totally static.

Let's think of theory as an engine instead.

Okay, I'm tracking.

Your independent variable is the gasoline.

Your dependent variable is the wheels turning.

Oh, I see.

The theory is the internal combustion engine that explains the exact physical mechanism of how that gasoline actually forces those wheels to turn.

It gives you the rationale for your expectation.

Okay, I like the engine analogy much better than the rainbow, but wait, the book mentions that theories vary wildly in their breadth of coverage.

Right, based on Neumann's work.

Yeah, it outlines micro, meso, and macro levels.

So if a theory is just an engine connecting two variables, why does the size of the engine even matter?

Well, the size of the engine dictates the scope of the reality you're trying to explain.

So micro -level theories are limited to really tiny slices of time, space, or just small numbers of people.

Like what?

The book uses the example of a theory of facework.

It explains the very specific psychological rituals people engage in during brief face -to -face interactions to save face or avoid embarrassment.

So it only explains a micro interaction, like a two -minute conversation.

Right, it's a small, highly specific engine, but then you scale up to meso -level theories.

Meso meaning middle.

These link the micro and the macro.

They're theories of organizations, communities, or social movements.

Oh, okay, like a theory of control within a corporate hierarchy or something.

Spot on, it explains how a mid -sized group operates.

And finally, you have macro -level theories.

The big ones.

The massive ones.

These attempt to explain huge aggregates, social institutions, cultural systems, or whole societies.

Like a theory of social stratification, maybe, where you try to explain how an entire society's economic surplus affects its class structure.

Exactly, but here's the key.

Regardless of whether you're explaining a quick conversation or the economic structure of a nation, the core job of the theory remains the same.

It bridges variables.

It bridges variables.

Which means before we can build any kind of theoretical engine, we need to know what our parts are.

Like we need to clearly define these variables.

Because you cannot test a causal claim if you don't know what you're actually measuring.

Exactly, and this is where terminology trips up a lot of students, I think.

Oh, all the time.

But at its core, a variable is simply a characteristic or attribute of an individual or an organization that can be measured or observed.

And crucially, that varies among the people being studied.

Right, if everyone in your study is exactly 20 years old, age is not a variable.

Because it doesn't vary, it's just a constant.

Precisely.

So let's run through the specific types of variables in quantitative research.

The primary two are the independent variable and the dependent variable.

Right, the classics.

So the independent variable is the cause.

It's the thing influencing the outcome.

And in a true experiment, it's the specific thing the researcher's physically manipulating.

Right, and then the dependent variable is the outcome.

I always remember it because it literally depends on the independent variable.

That's the best way to remember it.

And the textbook uses this great red wine study to illustrate this.

Oh yeah, let's talk about the wine.

So imagine an eight week medical study.

The researcher assigns one group of adults to drink one moderate glass of red wine every single day.

Nice study to be a part of.

Right, that is the experimental group.

Now another group is told to abstain completely.

No wine, that's the control group.

Okay.

Since the researcher is systematically manipulating the wine consumption,

moderate red wine consumption is our independent variable.

And what are they looking for?

Like heart attack incidents,

plaque formation in the arteries,

strokes?

Exactly, those are the dependent variables.

The physical state of the patient's heart depends on the amount of wine they were assigned to drink.

Okay, but what if you aren't doing an experiment?

What if you're just, I don't know, mailing out a survey to ask people about their normal weekend habits?

Ah, then the terminology actually shifts to reflect that lack of control.

Because you aren't manipulating anything.

Exactly.

In survey method studies, researchers can't systematically manipulate who drinks what.

They just observe what's already happening.

Okay, so what do we call them then?

We call the suspected cause a predictor variable

and the suspected outcome an outcome variable.

Predictor and outcome, got it.

So they function just like independent and dependent variables.

Yes, but the new name signal to the reader that the researcher didn't control the inputs.

Okay, those are the standard parts of the engine.

But there are three incredibly tricky variables that sneak into quantitative research and we really need to lock these down for you.

Yes, let's keep the red wine flowing to help memorize them.

Perfect.

First up is the mediating or intervening variable.

So a mediating variable stands directly between the independent and dependent variable.

It is the actual mechanism transmitting the effect.

So in our wine study, the mediating variable would be like the polyphenol compounds found in the grape skins, right?

Exactly, it isn't the physical act of holding a wine glass that helps your heart.

Right.

It is the polyphenols entering your bloodstream that exert the actual physical health benefit.

The mediating variable explains the how.

It's the chemical middleman.

Okay, next is the moderating variable.

I kind of think of this as the boundary condition.

That's a great way to frame it.

It affects the direction or the strength of the relationship between your main variables.

And moderating variables are very often demographic traits, aren't they?

Very often, yes.

Let's say this study proves red wine reduces heart attacks.

But a moderating variable might reveal that this relationship is drastically stronger for people over the age of 60.

Oh, and maybe almost nonexistent for people in their 20s.

Exactly.

The age of the participant moderates the effect of the wine.

Got it.

And finally, the absolute villain of the research world,

the confounding variable.

The sneaky third variable.

A confounding variable is related to both the independent and dependent variables.

And if you fail to measure it, it will completely ruin the validity of your study.

It will destroy it.

So imagine you run the study and the data clearly shows that people who drink a glass of wine every night have historically low rates of heart attacks.

Okay.

You publish a paper claiming wine is a miracle cure.

But what if the specific type of person who carefully drinks one moderate glass of wine with dinner also happens to be the type of person who does 45 minutes of rigorous aerobic exercise every single morning?

Oh, wow.

Right.

What if it's the exercise preventing the heart attacks and the wine is doing absolutely nothing?

If you didn't think to measure their exercise habits, your causal claim is totally garbage.

You measure the wrong engine.

Exactly.

Which really highlights why measuring variables accurately is paramount.

It is.

And the textbook briefly outlines how researchers score these variables using different scales.

They generally fall into two buckets.

Categorical or continuous.

Okay, categorical puts people into distinct groups, right?

Yes.

You might use a nominal scale for something like education level, high school, bachelor's, master's, or an ordinal scale where the categories have a clear rank order.

And then continuous scales are more like sliding sliders than buckets.

They use intervals or ratios.

Like a standard survey asking you to rate your agreement with the statement from one to five.

Strongly disagree to strongly agree.

Right, so now we have our parts.

We have our independent gas, our dependent wheels, our mediating mechanisms, and we know how to measure them all.

So now we need to prove that one actually causes the other.

We have to put quantitative theory into practice.

And establishing causality requires strict rules about time, doesn't it?

Oh, absolutely.

Temporal order is the golden rule of quantitative research.

Variable X must occur before variable Y in time.

Cause before effect.

Exactly.

Because of this strict adherence to time, quantitative researchers always conceptualize, read, and write their models from left to right.

So the cause is always on the left, moving across the page to the effect on the right.

Always.

But practically speaking, if I'm a student trying to write a research proposal, how do I actually formulate these theories on the page?

Do I just write a paragraph?

Well, the textbook highlights three primary ways to state a quantitative theory.

Okay, what's the first?

The first is as interconnected hypotheses.

A researcher might write a series of statements that chain together.

The text uses Hopkins' 15 hypotheses on influence processes as an example.

Oh, that's the one where everything triggers the next thing, right?

Like the higher a person's rank, the greater their centrality.

The greater their centrality, the greater their observability.

Exactly.

It reads like a literal chain reaction of variables.

Okay, what's the second way?

Second way is through if -then logic statements.

This is highly common.

Hallman's theory of interaction is the textbook's example here.

How does that one go?

It says if the frequency of interaction between two people increases, then their degree of liking for one another will increase.

Oh, I see.

It establishes a very clear testable conditional mechanism.

Right.

And the third way, and this is honestly where students usually panic, is visual causal models.

Oh man, the textbook is full of these intense diagrams with boxes and arrows pointing everywhere.

It's overwhelming.

How does a student translate a massive visual flow chart into something they can actually understand?

The trick is to stop looking at it as a static picture.

You have to start reading it as a strict timeline of a person's life, reading left to right.

Okay, so the arrows indicate the flow of time and influence.

Yes.

And you'll also notice valence signs on those arrows.

A plus sign means a positive relationship.

As the variable on the left goes up, the variable on the right goes up.

And a minus sign means a negative relationship.

As one goes up, the other goes down.

Precisely.

Let's apply that timeline trick to the scariest diagram in the chapter.

Yeah.

Jung -Nickel's model on faculty scholarly performance.

Oh yeah, that one is a massive web.

It is.

But if we read it as a timeline, we just start on the far left.

And to read complex models like that, you need to understand two concepts, exogenous and endogenous variables.

Okay, prick those down for me.

Exogenous variables sit on the extreme far left of the timeline.

They are not caused by anything else in the model.

So in Jung -Nickel's study on what makes college faculty publish research, things like demographic variables or prior research training are exogenous.

Because your prior training from 10 years ago isn't caused by your current workload today.

Right, it's just a historical fact you bring with you into the present environment.

Right.

As we move to the right on the timeline, we hit endogenous variables.

These are includes by other variables in the model.

So Jung -Nickel lists workload and pressure to conduct research as endogenous.

Yes.

They are caused by your current environment and in turn, they push forward pointing rightward to influence the final dependent variable at the very end of the timeline.

Which is scholarly performance.

Exactly.

And this whole left to right cause and effect process is called the deductive approach.

It is entirely top down.

You start way up high with a broad theory.

You move down to formulate specific hypotheses.

You define the variables to test them.

And finally, at the very bottom, you measure the data to see if the theory holds up.

It's a highly scripted process.

The book even provides a four step script for writing this out.

Oh right, example 3 .1.

It looks at Crutchfield's dissertation on nursing educators.

She used social learning theory to predict scholarly productivity.

She didn't just guess.

She named the established theory, stated the central hypotheses,

explained who would use the theory before and then systematically applied it to her specific nursing variables using strict if then deductive logic.

Okay, so that is the rigid mathematical top down world of quantitative theory.

But what happens if you are studying something completely unprecedented?

What if you don't know enough about the topic to even guess at an if then statement?

Well, that brings us to the exact opposite methodology.

Qualitative theory use.

The inductive flip.

The inductive flip.

In quantitative research, the theory is a rigid engine locked in at the very beginning to be tested.

Qualitative research throws that out the window entirely.

Because it's exploratory.

Exactly.

Because of that, the use of theory is far more fluid.

The textbook outlines four distinct variations of how a qualitative researcher might interact with theory.

Let's go through them.

Variation one is deductive explanation.

Right, this is the closest to what we just talked about, but it's used much more loosely.

Like an ethnographer taking a broad cultural theme, maybe the concept of kinship or social control, and using it as a general starting point to study a remote group of people just to see if the theme applies.

Then variation two is where things get philosophically interesting.

This is theory as a theoretical standpoint or lens.

Starting in the 1980s, qualitative research evolved to use specific theoretical lenses designed to lift marginalized voices or challenge societal power structures.

I wanna pause here and clarify something really important for you, the listener.

The textbook lists several highly specific social justice frameworks here,

like feminist perspectives, critical theory, queer theory, racialized discourses, and disability inquiry.

We are mentioning these strictly to impartially report what is in your textbook.

We aren't endorsing any specific viewpoint, but rather we are explaining how these lenses function as methodological tools that fundamentally alter research design.

And that's a vital distinction, because using one of these lenses completely changes the physical actions of the researcher.

How so?

It dictates the types of questions asked, how data is collected, and how the researcher positions themselves relative to the subjects.

For example, look at disability inquiry.

A traditional medical model might look at a physical disability and ask, how do we fix the biological deficit in this individual?

But if a researcher applies a disability inquiry lens, the entire focus shifts.

They look at the society, not the biology.

So the research question becomes, what are the sociocultural barriers and environmental structures that are disabling this individual?

Exactly.

The lens completely changes the data you look for.

Another example the book gives is a feminist perspective, which centers women's diverse situations and interrogates oppressive structures.

Yes.

So if you apply that lens to a study on corporate leadership, you aren't just counting how many female CEOs exist.

No, not at all.

You'd be designing qualitative interview questions to uncover the specific systemic power dynamics that prevented women from reaching the boardroom in the first place.

The lens literally directs the spotlight.

It does.

Now, variation three is the true inductive process, theoretical endpoint.

Okay, meaning the qualitative study results in a theory.

Yes.

It is the literal opposite of the quantitative approach.

Right, so if the quantitative approach was top down, this is bottom up.

You don't start with a theory to test.

No, you start at the ground level, gathering detailed, messy stories and interviews from participants.

Then you read through those transcripts and ask open -ended questions.

Right, you start clumping the data into themes.

You look for broad patterns connecting those themes.

And finally, after all the data is sorted, you climb to the very top and build a generalized model or a grounded theory.

Exactly.

The theory is the destination, not the starting line.

And the textbook offers a brilliant contrast to show how the placement of the theory dictates the type of study.

Oh, example 3 .2, Mergia's study on Hispanic and Native American college integration.

Yes, they used an existing model, the Tinto model of social integration at the very beginning of the paper as a lens.

And then they spent the rest of the study modifying it based on their qualitative findings.

Compare that to example 3 .3, Creswell and Brown's study on academic department shares.

Yeah.

They didn't start with any model at all.

Right, they sat down and conducted 33 in -depth interviews first.

Exactly.

They analyzed the raw data, found the patterns, and placed their visual theoretical model at the very end of the paper.

The theory was the final product of the research.

Which is so cool.

And just to round out the list, variation four is no explicit theory.

Pure description.

In methods like phenomenology, the researcher deliberately tries to set aside all theoretical frameworks.

Just to capture the raw detailed essence of a human experience without analyzing it through any specific engine or lens.

Precisely.

So we have the strict left to right deductive quantitative approach and the bottom up inductive qualitative approach.

But modern research is rarely that clean, right?

Often we want the hard numbers and the rich personal stories.

Which brings us to blending the two together in mixed methods theory use.

Mixed methods requires a framework robust enough to handle surveys and interviews simultaneously.

And the textbook categorizes these frameworks into two distinct types.

First up is discipline based theories.

These are drawn directly from established social, behavioral, or health sciences.

So they identify specific variables to measure, but they also leave wide open spaces for exploring lived experiences.

Exactly.

The text highlights Kenneth's study on chronic pain management as the perfect discipline based example.

They used Rosenbaum's psychological theory of self control as their framework, right?

Yes.

They needed numbers.

So they gathered quantitative survey data measuring patients self control scales.

But numbers don't tell the whole story of pain.

Right.

So they also conducted qualitative, open -ended interviews with the patients.

Exactly.

The single psychological theory informed both the survey creation and the interview questions.

That makes a lot of sense.

And the second type of mixed methods theory is social justice oriented.

This builds directly on those transformative lenses we discussed a minute ago.

The text uses Hodgkin's study on women's social capital in Australia.

Right.

She applied a feminist emancipatory lens to the entire mixed methods project.

She used the quantitative data to map the landscape proving with hard numbers that there was a measurable difference in social capital between men and women.

But then she used the qualitative interviews to figure out the why.

She sat down with the women and discovered themes that numbers could never capture.

Like the desire to avoid social isolation or the intense pressure to be perceived as a good mother.

So the numbers showed the gap, but the narratives explained the gap.

Beautifully stated.

Now a perceptive student might notice a potential point of confusion here.

What's that?

Well, in the very first chapter of this textbook, you learned about philosophical worldviews, like the constructivist worldview or the transformative worldview.

Oh, right.

So if I decide I wanna use a feminist emancipatory lens because I believe it's important to highlight marginalized voices,

isn't that just my philosophical worldview?

How does the textbook distinguish a worldview from a theory?

It is a really vital distinction.

And the textbook relies on Crotty's four level model to untangle it.

To understand Crotty's model, don't think about a flow chart.

I want you to think about the process of building a house, moving from abstract ideas down to physical tools.

I love this analogy.

Walk me through the four levels.

Level one at the very top is your paradigm worldview.

These are your most abstract philosophical beliefs about reality.

So it's the architect's belief that a house should be in harmony with nature.

In research, a transformative worldview is the abstract belief that research should challenge oppression.

Okay, worldview is the abstract philosophy.

Level two is the theoretical lens.

We move from philosophy to literature.

You choose a specific established theory drawn from academic literature, like a specific feminist social justice theory that aligns with your worldview.

So this is the blueprint for the house.

It's more concrete than a philosophy, but you still can't live in it.

Brilliant, yes.

Which brings us to level three, the methodological approach.

Right, your blueprint dictates your approach.

If your theoretical lens requires understanding both broad structural inequalities and deeply personal experiences, your approach must be mixed methods.

This is deciding that your house needs both wood framing and poured concrete.

Which leads to the final level, level four, methods.

The literal tools in your hands, because you chose mixed methods, your actual data collection methods will be a combination of survey instruments and face -to -face interview protocols.

So Crotty's model shows that theory level two is the crucial translation step.

It is.

It is the actionable blueprint that takes your abstract worldview and turns it into physical, on -the -ground data collection.

To wrap this all together, the book gives us example 3 .4 to show how a single theory completely dominates a mixed method study from top to bottom.

Yes, Clark and Plano Clark conducted a study on grit and career success.

They selected positive psychology as their discipline -based theoretical framework.

And that single choice dictated absolutely everything that followed.

Everything.

Positive psychology framed the initial design.

It dictated the exact quantitative survey tools they use to measure the grit of 423 adults.

Wow.

It guided the criteria they used to select five specific individuals for deep -dive qualitative interviews.

It literally shaped the phrasing of the interview questions.

And finally, it provided the lens through which they interpreted the final converged data.

Exactly.

The theory wasn't just an introduction paragraph.

It was the steel cable running through every single phase of the project.

Which brings us back to the core message for anyone tackling the syllabus.

Theory is not an academic hurdle designed to bore you.

Not at all.

It is the fundamental mechanical logic of your research design.

Whether you are building an engine to test a hypothesis deductively from left to right.

Or you're gathering messy data to build a grounded theory inductively from the bottom up.

Theory is the mechanism that gives your data actual meaning.

Absolutely.

You survived the deep -dive into chapter three.

We know it is a dense chapter.

But if you can internalize the difference between top -down deductive and bottom -up inductive, and if you can memorize those tricky variables, you are gonna absolutely crush this section of your course.

Just remember to approach your visual models as a timeline flowing left to right, and always interrogate the underlying rationale behind any assumed relationship between two variables.

Which leaves us with a final thought for you to mull over as you close your textbook today.

The next time you are scrolling through your phone and you see a flashy clickbait news headline claiming that variable X absolutely causes variable Y, I want you to pause.

Put your researcher hat on.

Yes, put your researcher hat on and ask yourself two questions.

First, what is the unmeasured confounding variable they completely missed that is actually causing the outcome?

And second, what invisible theoretical engine are they using to try and bridge that gap?

Thinking critically about cause and effect is your best defense against bad information.

Keep interrogating those variables.

Keep looking for the structural cables, holding the claims together.

Good luck with your studies and a warm thank you from the last -minute lecture team.