Chapter 13: The Descriptive Research Strategy

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.

Welcome back to the Deep Dive, where we take your sources, slice through the noise and really pull out what matters, the key insights, the context you need.

Today, we're tackling something pretty fundamental in research methods.

It's the descriptive research strategy.

We're using a chapter from a Research Methods for the Behavioral Sciences textbook as our guide.

And our mission really is to walk through this whole chapter together, you know, piece by piece, and get a solid handle on how researchers simply describe the world.

Not necessarily why things happen, but just what's going on.

Exactly.

That's the key distinction, isn't it?

Descriptive research, it isn't about explaining relationships or finding causes.

It's not asking why.

The whole point is just capturing a snapshot, describing variables exactly as they are, naturally, painting that picture.

And this chapter highlights the three main ways researchers do that, observational designs, survey designs, and case study designs.

Right.

And when you stop and think about it, those methods are behind so many of the facts you bump into every day, you know, like X percent of people don't get enough sleep, or figures on obesity rates, average student debt, how much time we spend looking at screens, that kind of thing.

Classic descriptive findings, just telling us the current state of play for a variable.

OK, so let's dig in.

Let's unpack this, starting right at the beginning with the core idea the chapter introduces.

So fundamentally, like you said, it's about describing variables as they naturally exist.

The chapter really hammers this home.

It's not about finding relationships or explaining causes.

Precisely.

You hear about experimental studies, correlational studies, quasi -experimental.

Those are all geared towards understanding connections, right?

Influence, comparisons, cause and effect.

They're asking is or how or why things are linked.

But descriptive research just steps back from all that.

It does.

It essentially says, hang on, before we get into the why, let's just get a really clear picture of what we're even looking at.

And that's incredibly valuable, especially

when you're just starting to explore a topic.

You need that clear description first.

It's about capturing things as they naturally unfold.

And the main tools for that, as the tactor lays out, are those three designs.

Watching people that's observational, asking people that surveys, and then really zooming in on one person or situation, the case study.

Let's kick off with the first one, observational research.

The chapter goes into quite a bit of detail here.

It's all about systematically watching and recording behavior.

Yeah, and it's worth pausing on a point the chapter makes.

Observing behavior, well, that happens in lots of research, right, even experiments.

But when we talk about the observational research design, the observation is the main event.

It's used only to describe that behavior.

That's the key difference.

Okay, but just watching people.

That sounds like it could get tricky.

It definitely can.

Two big challenges pop up immediately.

The first is observer influence.

Ah, like people acting differently because they know they're being watched.

Exactly.

You get things like demand characteristics participants trying to guess what you want or just plain reactivity, changing behavior because someone's there.

So how do researchers get around that?

Well, two main strategies mentioned.

One is concealment, hiding the observer.

Ethically, that's generally only okay for public behavior in public places, you know.

Right.

You can't just hide in someone's living room.

No, definitely not.

The other way is habituation.

Basically, the observer hangs around long enough that people just get used to them and ideally go back to behaving normally.

Think wildlife documentaries.

The animals eventually ignore the cameras.

Okay, so that deals with people changing their behavior.

What's the second big challenge?

Subjectivity.

Observation isn't always clear cut.

Was that push playful or aggressive?

Different observers might see it differently.

The interpretation problem.

Precisely.

So researchers need ways to make it objective.

They use really well -defined behavior categories like a checklist of specific actions.

They use observers who are thoroughly trained.

And crucially, they often use multiple observers watching the same thing and then calculate inter -rater reliability.

To see how much they agree.

Exactly.

High agreement gives you confidence that it's not just one person's opinion.

So you've got your trained observers, your checklist, but how do you actually turn that watching into, well, into data?

Numbers you can work with.

Right.

Chronification.

The chapter outlines three main techniques for this.

First, there's the frequency method.

Super simple.

You just count how many times a specific behavior occurs within a certain period.

Like a kid throws a toy five times in ten minutes.

Yep.

Frequency is five.

Second is the duration method.

You measure how long a behavior lasts.

So the kid played alone for, say, 18 minutes straight.

That's duration.

And third, there's the interval method.

This one's a bit different.

You break your observation time into smaller intervals, maybe a minute each.

Then for each interval, you just record whether the behavior happened at all during that short window.

Not how many times or for how long.

Just yes or no for that interval.

Exactly.

It's really useful when behaviors are continuous or happen in rapid bursts.

It kind of balances frequency and duration.

The book uses bird songs as an example.

Some long, some short interval catches the presence over time.

Okay, that makes sense.

But what if you're watching something really complex, like a No,

you definitely can't.

And that's where sampling techniques become absolutely essential.

You have to sample the behaviors or the time.

So one way is time sampling.

You observe for a set period, say one minute, then you take the next minute to record everything you saw, then you observe again.

It's like taking snapshots in time.

Observe, record, observe, record.

Right.

Then there's event sampling.

Maybe you focus only on sharing behavior for the first five minutes.

Then you switch and only look for conflict behavior for the next five minutes.

You're sampling specific events or behaviors sequentially, focusing on one thing at a time.

Exactly.

And third is individual sampling.

You watch one child for 10 minutes, then switch your focus entirely to another child for the next 10 minutes and so on.

Gotcha.

And I guess nowadays you could also just video record the whole thing and analyze it later using these methods.

Absolutely.

Recording gives you the luxury of rewatching and applying different sampling or quantification techniques after the fact.

It's a huge help.

Now here's a really interesting bit.

What if the behavior you want to describe didn't happen live?

What if it's in the past or maybe in a movie or a book?

Great question.

That brings us to content analysis and archival research, which the chapter also covers under the observational umbrella in a way.

Content analysis is basically applying those same systematic observation techniques, defining categories,

quantifying, using multiple raters for reliability, but you're applying them to media.

Like books, TV shows, ads.

Exactly.

The chapter mentions a cool study looking at older adults in Super Bowl commercials.

They defined categories for older adult presence,

watched commercials from several years, and quantified how often they appeared.

And what did they find?

Just out of curiosity.

Interestingly, older adults showed up in over 30 % of the commercials they analyzed, which was actually more than their proportion in the general population at the time.

A neat descriptive finding from analyzing media content.

Huh.

Okay, and archival research.

Sounds like digging through old files.

Pretty much.

It involves analyzing existing historical records archives.

Think birth certificates, marriage licenses, public health records, crime reports, diaries.

Anything that documents past behaviors or events.

The example given is a study that looked at marriage records and found people were statistically more likely to marry someone with a similar first or last name than you'd expect by chance alone, using archives to describe a pattern from the past.

So even for past stuff or media, you're still using those core observational principles, categories, quantification, reliability checks.

That's the key.

The methods are similar, just applied to different kinds of behavioral records.

Okay, let's zoom back out a bit on observation.

The chapter talks about different types based on the setting and the observer's role.

Right.

Three main types are discussed.

First is naturalistic observation.

Like Jane Goodall with the chimps.

Exactly.

That's the classic example.

The researcher observes behavior in its natural setting, trying to be as unobtrusive as possible non -participant.

Another example mentioned was video recording couples interacting naturally at home.

What are the pros and cons there?

Big pro.

High external validity.

You're seeing behavior as it really happens in the real world.

It's essential for things you can't ethically manipulate, like studying natural parenting styles.

It can be incredibly time -consuming just waiting for behaviors, and you still have those potential problems of observer influence, even if you try to hide, and observer subjectivity.

Okay, type two.

Participant observation.

Here, the researcher actually joins and becomes part of the group they're studying.

Whoa, like going undercover?

Sometimes, yeah.

It's necessary when you can't observe from the outside, maybe studying a very private community, a cult, a gang.

The chapter mentions Rosenhan's famous controversial study where researchers fake symptoms to get into psychiatric hospitals.

That sounds intense and risky.

It can be.

Strengths are gaining access you couldn't get otherwise, getting a unique insider perspective.

Weaknesses include the potential danger, the very real possibility the researcher's presence changes the group's behavior, and the huge challenge of staying objective when you're involved.

Makes sense.

And the third type.

Contrived observation, or sometimes called structured observation.

Here, the researcher sets up a specific situation designed to elicit the behavior they want to see, making it happen faster than waiting for it naturally.

Like bringing kids into a lab playroom.

Exactly.

Or setting up a specific problem for them to solve, like Piaget did.

It could also be in the field, like putting out a specific type of bird feeder to observe feeding behaviors.

Benefit being speed, I guess.

Definitely.

You don't have to wait around hoping the behavior occurs.

The downside, of course, is that the setting isn't fully natural, so the behavior might be slightly less authentic.

So wrapping up observational research overall,

what are the big takeaways?

Well, the major strengths are that you're observing actual behavior.

You can get high external validity, especially in naturalistic settings, and it's flexible.

Weaknesses include potential ethical issues around privacy, spying, dealing with observer influence and subjectivity, and the big one for all descriptive research.

It describes what, but doesn't explain why.

Okay, that's a fantastic overview of watching.

Now, let's completely switch gears to the next big descriptive tool.

Survey research.

Asking people.

Right.

The survey research design, its core purpose is pretty straightforward.

It's an efficient way to gather information about people's attitudes, opinions, personal characteristics, past behaviors,

things that are hard or impossible to observe directly.

You just ask them.

Just ask them.

But again, the chapter stresses a key point.

Just using a survey doesn't automatically make it a survey research design.

Lots of studies use questionnaires.

So what makes it this specific design?

The goal.

If the sole purpose of using that survey is to describe the characteristics or opinions of the group you surveyed, that's a survey research design.

Think basic customer satisfaction surveys, political polls describing voter intentions.

Gotcha.

So if you're building a survey for descriptive research, what are the critical things to get right?

The chapter breaks it down into four key issues you really need to think through carefully.

One,

question development, crafting the actual questions.

Two, survey construction, putting those questions together logically.

Three, sample selection, deciding who you're going to ask.

And four, administration method, how you'll actually get the survey to them.

All right.

Let's tackle those one by one.

First, the questions themselves.

What kinds are there?

The chapter highlights three main types.

First, open -ended questions.

Where people just write or say whatever they want.

Exactly.

What are your thoughts on X?

Strengths or flexibility.

You can get really rich, unexpected answers.

Weaknesses.

They're a nightmare to analyze statistically.

Comparing answers is tough, and it depends on how well people express themselves.

Okay.

What's the alternative?

Restricted questions.

These give people limited options like multiple choice, checklists, or true -false.

Much easier to count up the answers.

Way easier.

You get clear percentages, proportions.

You can analyze them easily.

You could even add an other please specify line to catch things you missed.

Main weakness is, well, you're restricting their answers to your choices.

And the third type.

Rating scale questions.

These ask participants to pick a numerical value on a scale to indicate degree or amount.

Like the classic, on a scale of one to five, how satisfied are you?

Exactly.

The Likert scale,

strongly agree to strongly disagree, is a super common example.

The chapter mentions using anchors labels for the points or ends of the scale to make the meaning clear.

What's good about those?

They give you numerical data that's easy to work with, statistically calculate averages, totals, et cetera.

Participants usually find them easy and quick to answer.

You can gather a lot of data efficiently.

Any downsides?

One potential issue is the response set.

People might just fall into a pattern, like always picking the middle number or always agreeing.

Researchers try to counter this by mixing up positively and negatively worded statements.

Okay.

So you've crafted your different types of questions.

How do you assemble them into a good survey instrument?

There are some general guidelines.

Put demographic questions, age, gender, et cetera, near the end.

They can feel a bit personal upfront.

Right.

Ease people in.

Exactly.

If you have sensitive questions, maybe put them in the middle once people are already committed.

Group similar questions together, especially ones using the same format, like all your rating scales.

Keep the layout clean, uncluttered, and use simple language your participants will understand.

Makes sense.

Don't make it hard work for them.

Right.

The easier it is, the better the data you'll likely get.

Okay.

Survey built now.

Who do you give it to?

Selecting participants.

This is huge, especially for external validity, being able to generalize your findings.

Your sample has to be representative of the group you want to describe.

So if you're serving parents about childcare, you need to survey parents.

Seems obvious, but yeah.

And if you want to describe a broad population, like all adults in the city, your sample can't be restricted to just, say, university students.

It needs to reflect the diversity of that broader group.

Sometimes researchers use specialized lists or companies to reach specific target populations.

All right.

Final piece.

Getting the survey to those people.

Administration methods.

Yep.

The chapter covers four main ways, each with its own pros and cons.

And we need to think about things like non -response bias and interviewer bias here.

Non -response bias being the people who don't answer are different from those who do.

Exactly.

Which can skew your results.

And interviewer bias is when the person giving the survey influences the answers, even unintentionally.

Okay.

So what are the four methods?

First, internet surveys.

Hugely common now.

Pro.

Cheap.

Fast for large numbers.

Easy to reach specific online groups.

Flexible with things like skip logic.

Cons.

Sample might not represent everyone.

Digital divide, self -selection.

Non -response bias is still an issue, and sometimes hard to know exactly who filled it out.

Okay.

Method two.

Mail surveys.

Sending them through the post.

School pros.

People can do it on their own time.

Feels anonymous.

Non -threatening.

Cons must be the response rate, right?

That's the killer.

Often really low, like 10, 20 percent.

That means potentially huge non -response bias.

Only the really interested or passionate people might bother sending it back.

Plus, it costs money for printing and postage.

The chapter suggests ways to boost rates, like a good cover letter, maybe a small incentive, reminders.

Method three.

Telephone surveys.

Calling people up.

I feel like I get fewer of those now.

Maybe, but they're still used.

Pros.

Researcher can do it from an office.

Participant is at home.

Cons.

Very time -consuming.

It's one -on -one, so you need lots of interviewers for big samples.

And interviewer bias is a risk -toner's voice.

How you read the question.

Keep questions short.

Practice reading neutrally.

Right.

Don't lead the witness.

Exactly.

The last method.

In -person surveys, or interviews, face -to -face.

Right.

Pros.

If you survey a captive audience, like a classroom, you get near a hundred percent response.

Interviews allow follow -up questions, getting deeper info.

Good for people who can't read.

Individual interviews are super time -consuming.

Interviewer bias is a big risk.

Smiling, nodding can subtly influence answers.

And you're still relying on self -report, which might not be totally accurate.

Okay, so summing up surveys.

Strengths.

Flexible, efficient for gathering lots of data on various topics.

Weaknesses.

Non -response bias is often a major headache.

Analyzing open -ended answers is tricky.

And the core limitation is relying on self -report people might not remember accurately or might want to look good.

Perfect.

We've covered watching and asking.

Now for the third descriptive strategy.

The case study design.

Really zoning in on one individual.

This is quite different from

what we call the nomothetic approach, which seeks general laws.

Case studies are idiographic, focused intensely on understanding a single individual in great detail.

Very common in clinical psychology.

So it's more than just an anecdote.

Oh, much more.

It involves collecting extensive information interviews, observations, records, tests about that one person.

If there's no treatment involved, it might be called a case history.

But the key for this design is that the goal remains description.

You're describing that individual case, not necessarily proving a cause from that one instance.

So when does it make sense to put all that effort into just one person?

Why do a case study?

The chapter highlights a couple of really important situations.

First, for studying rare phenomena or unusual clinical cases.

Things where you just can't get a large group together.

Exactly.

Like studying someone with a very rare condition.

Or dissociative identity disorder.

Multiple personalities, as mentioned in the text.

Or critically, understanding the effects of unique brain injuries.

Ah, like the famous case of H .M.

Precisely.

Patient H .M., Henry Molaison.

His surgery left him unable to form new long -term memories.

Studying him in exhaustive detail completely revolutionized our understanding of memory systems in the brain.

You simply couldn't ethically replicate that kind of situation.

You learn from the unique cases that arise.

Wow.

Okay, what else are they good for?

They're also used to demonstrate new therapy methods or applications.

A detailed case study can illustrate how a new technique was applied, and how it seemed to work for one specific person.

Like a proof of concept.

Kind of, yeah.

The chapter mentions a study describing how adding exposure therapy helped one specific person with anorexia manage severe food anxiety.

It shows the potential application.

So this intense focus.

What are the main upsides and downsides?

Strengths are definitely the incredible level of detail.

You can uncover nuances, complexities, and potential new hypotheses you'd miss in a group average.

They're great for identifying exceptions to general rules.

And they feel very real.

Right.

Absolutely.

They can be very powerful, convincing, memorable.

The chapter compares hearing accident statistics versus the story of one specific accident.

The single story often has more impact.

They have a realness, especially in clinical settings.

Okay, but what are the weaknesses?

The biggest one is limited generalization or external validity.

Because it's just one unique person or situation, you don't know if the findings apply to anyone else.

It could just be that person.

Exactly.

They also lack internal validity.

You can describe what happened, but the case study alone can't definitively prove why it happened.

And bias is a risk.

Researchers might selectively report successful cases,

their own interpretations are subjective, and the client's self -report might be biased or inaccurate.

So tricky to draw broad conclusions.

Very.

Though the chapter does note that replication finding multiple similar case studies can build more confidence, but the inherent limitation in generalizing from one or a few individuals remains.

Wow.

I think we've actually walked through the whole chapter's territory on descriptive research.

We hit the core idea just describing variables.

And the three main designs, observational research, watching behavior, survey research, asking people, and case studies.

That deep dive into one individual.

Yeah, we looked at how you collect and quantify data, especially for observation and surveys.

Went through some key examples from the text for each method and really talked through the pros and cons, the strengths and weaknesses.

The core message repeated throughout is that this strategy gives us that snapshot.

What is the situation?

It's foundational, even though it doesn't tell us the why.

And understanding all this, well, it gives you, the listener, a completely new way to look at information you see every day, doesn't it?

You see a news report with the statistic or a claim about public opinion or how people behave.

Now you can start thinking, okay, how did they find that out?

Was it observation, a survey, maybe even based on case studies?

Exactly.

Who did they watch?

Who did they ask?

And knowing the weaknesses we discussed lets you evaluate it critically.

Like, hmm, that was a mail survey with only a 15 % response rate.

Who didn't answer?

And how might that affect this finding?

Or this was participant observation.

Could the researcher's presence have changed things?

So next time you see a headline that starts,

a survey finds or observation shows, you've got a much better toolkit for understanding what's really behind that claim.

You can start spotting the potential strengths, but also the limitations yourself.

Which leads to a really interesting question to maybe ponder, thinking about those different strengths and weaknesses we just covered for observations, surveys, and case studies.

If you wanted to describe, say, how people actually use a new public park, or maybe people's opinions on building a new community center,

or perhaps understand a very, very rare fear someone has, which descriptive design do you think would be the best starting point for each of those questions and why?

Something to think about.

We really hope this deep dive into descriptive research has armed you with some valuable insights into how we map out the what's of the world around us.