Transcription
The video lecture is on module 2 and 3. We are talking about studying human behavior. So, the foundation of this unit is all about research. Why do we need to study human behavior? Can't we just look at human behavior and state what's obvious and call it a day? Why do we need to actually research human behavior? And that's what we're going to look at today. One, why? What's the importance of conducting research? And two, how do we conduct research?
So, psychology is a science. We are called, um, a social science because we use the scientific principles that you are familiar with here in your science classes, but we use them in the social setting. We find ways to control and manipulate and measure human behavior in order to draw conclusions about why we act the way we do, why we feel, think, um, why relationships turn out the way that they do, why friendships turn out the way that we do. Everything that we do in the field of psychology is based on science. How can we prove human behavior rather than just be a bunch of fluff and just say what we think?
So, um, one of the reasons why we need to do this is because of something called hindsight bias. I think you're going to experience this more than one time throughout this semester, um, throughout this entire school year. Hindsight bias is the "I knew it all along" phenomenon. It's this idea that when we are are faced with evidence or information, usually after the fact, we are very quick to say, "Well, oh yeah, I mean, I knew that that was going to work." So, take for example, you're watching a basketball game and the coach makes a really risky, um, play and it and it works out, and at the end of the game, you say, "Yeah, I knew that that was going to work. I knew that he would call the right play." Or take it in the opposite direction. Close game, the coach calls a really risky play, it doesn't work out, and you say at the end of the game, "Well, yeah, I knew it wasn't going to work." We tend to think after the fact that we would have known that all along. Say we are, um, choosing a number between one and 100, and we pick a number and a different number is chosen at the end. We say, "Oh, well, yeah, I knew that was the number that was going to be chosen." Well, if you knew it, why didn't you choose it first?
So, a lot of times we conduct research to avoid this hindsight bias, to avoid us saying, "Oh yeah, I knew it all along." Rather, we can prove what happens. We can prove why humans act and behave the way that they do. One experiment, for example, let's say that I was to independently give you, um, this piece of evidence that absence makes the heart grow fonder. Given that information about long-distance relationships, you might say, "Yeah, duh, I knew that." I might be able to come up with lots of examples as to why long-distance relationships work. Give another half of the class the results out of sight, out of mind, that long-distance relationships aren't good and it's going to deteriorate a relationship. We might have similar outcomes where people are saying, "Yeah, that's not surprising." Um, but the question is, which is true? And unless we conduct research, we can never know that answer.
So, a lot of times in psychology, you want to avoid that hindsight bias. When you've learned things in class, when you learn things, um, from reading research studies, recognize that hindsight bias does play a role. When we're faced with the outcome, it's easy to say that we would have known that the entire time.
So then, how do we study humans? Because they are very different than studying a rock or studying how what happens when two chemicals get put together, um, in one beaker. Studying humans is totally different. We're dealing with individuals. We're dealing with different personalities. We're dealing with different characteristics. So how do we go about studying humans?
So, so before we even can look into different research methods or ways of studying humans, we first have to have a theory. We have to have something that we're interested in. We have to have an idea about human behavior. Um, maybe, for example, I have a theory about whether or not long-distance relationships are good or bad. Maybe I have a theory that long-distance is good for relationships. That's that's my belief. My theory is that long-distance relationships are a good thing. Well, then my next step is I need to put that into a hypothesis. Is I need to put that into some measurable prediction. I need to say somehow that, you know, if you are in a long-distance relationship or if you are, um, miles apart, then your relationship is going to be stronger. Then you're going to have a better, um, relationship. You're going to have a better, um, more long-lasting relationship.
So, in creating an experiment, I have to first have a theory. I have to have a belief about something in human behavior. Then I need to put that into something that's testable. And ideally, in that thing that's testable, it needs to be falsifiable, meaning it needs to be able to be proven false just as much as it needs to be able to be proven true. It also needs to be able to be proven false. And so when I'm creating a hypothesis, make sure it's falsifiable. Am I able to find the opposite from what I think it's going to look like?
After I've done that, then I need to start to operationally define some of my variables. And operational definitions, I'm going to be hitting this home for the rest of the school year. So make sure in big bold letters you have: operational definitions are specific, they are measurable, and they allow for replication to occur. They are specific, they are measurable, and they allow for replication to occur.
By specific, I mean, let's say I'm studying whether or not long distance, um, makes a relationship stronger. I need to operationally define what is long distance. And when I mean be specific, I mean be so stupid specific, you are using numbers. So being specific as to what long distance is, I might say having a home residence that is 10 miles or more door-to-door. That is specific. Somebody else might operationally define long-distance relationship as, um, going to, um, living in a household that is in a different state. Maybe that's your operational definition. But it needs to be specific. Whatever it is, it needs to be specific to your research study, and you need to state that.
It also needs to be measurable. So when I'm saying I want to know, um, whether or not long-distance relationships is better for a relationship, I need to operationally define that measurable piece of what is better. What does it mean to have a better relationship? Am I going to operationally define better as, um, the number of months you date before you get engaged? Am I going to define better as the number of times you smile in a 10-minute period in the room together? Am I going to operationally define better as the number of phone calls that you have in an average day? Whatever it is, whatever I'm saying better is, when I operationally define it, I need to make sure it is specific and it is measurable. Give me numbers. Be specific.
And finally, this allows us to be able to replicate. This allows us to be able to say, um, we conducted this study in one instance. A study is not good unless it's replicated, unless we get the same results another time. We want to make sure our operational definitions are so specific that when somebody goes to replicate, they are replicating our exact experiment, meaning they are replicating it with the same operational definitions. They've operationally defined long distance as 10 miles door-to-door. They've operationally defined and measured a better relationship as the number of times you smile in a 10-minute interaction with one another. Whatever it is, you want to make sure that they are specific and measurable so that you can replicate. You can do this experiment again and find, hopefully, the same results more than once to prove that yes, this is in fact why.
Then you have to decide how you're going to study this. There are non-experimental methods and there are experimental methods. In our non-experimental methods, these methods merely describe behavior. These methods, um, will describe why, um, describe what behavior has occurred, but it isn't until we do an experimental method where we could explain why a behavior or explains where that behavior comes from. So, merely in non-experimental methods, I can describe the behavior. I can tell you that this person smiles this number of times in a 10-minute conversation. Nowhere can I explain why or make a connection that it is because of that relationship unless I do an experiment. I can't make a cause and effect. I can't explain, um, a behavior, explain why a behavior has happened.
So, our non-experimental methods again, where we describe behavior, we have a case study, naturalistic observation, survey, and correlational study. And then in our experimental method, the only method that we can use to explain cause and effect in in relationships, um, is the experimental method. Um, the experimental method is going to be a separate video lecture. So today, we are going to just focus on the non-experimental methods that again describe behavior. They don't explain why behavior is the way it is. They can't explicitly say a cause and effect that this behavior is due to this. They merely just will describe and present behavior to an individual.
So, the first method is a case study. A case study is an in-depth look at one person or one small group of individuals. Um, this oftentimes is done when you don't have a choice. Take the case of Phineas Gage that you see in this picture. Phineas Gage in the 1800s was working on the railroads, and there was some explosives that he was that they were utilizing. In one of the explosions, a tampering iron, which is a three-foot long nail, um, went under his left cheek and out his, um, the back of his frontal lobe. It shot about, you know, 50 feet behind him, and he he lived. He stood up and he he lived. He's still talked, um, but we look at some of the changes in behavior that occurred for him. We're going to talk about him a little bit more in depth during our, um, second unit, but he is a case of an individual where we would be forced to do a case study. There are not a lot of individuals who have had the blunt force trauma under their left cheek and out their frontal lobe in the exact same location that Phineas has had it. So we are forced to, if we're trying to see what that portion of the brain does, we are forced to do a case study on him where we'll gather as much information as we can about Phineas because we don't have anybody else to choose from. We don't have a whole group of people who have that same injury that we're able to gather that evidence from. Um, sometimes in psychology, we do case studies. Um, for example, a man named Jean Piaget studies something called cognitive development, or how thinking changes in children. In his early research, he did a case study on his own children and he looked at the thought patterns of just his children growing up in order to create his theory.
So, case studies are good in that we get a lot of information from them, but they're not so good in that we can't generalize the results of those. We can't say and determine that for sure there's a cause and effect of whatever it is that we are looking at is the reason why they have whatever it is the outcome that we're seeing. Or in the case of Phineas Gage, we can't say for sure that the damage to that portion of the brain causes this effect. We can describe and say, "Well, since this accident, he has had these types of outcomes," but we cannot draw a relationship saying that it is due to that.
The second technique is a naturalistic observation, and a naturalistic observation is where we are able to gather information and data, um, from individuals in their natural habitat. Probably the most common that you will, you know, you can come up with is animal research. You know, going out to animal habitats and looking at the behavior of those animals in their natural habitat. Um, in a naturalistic observation, you don't interact with the participants. You don't interact in that environment. Rather, you are merely on the outside observing, looking. So you might do a natural observation in the cafeteria, um, at the high school where you might go and you might, um, observe a certain behavior. Maybe you are looking at the snack buying among males, um, ages 15 to 17, when an authority figure is present, or you are looking at the snack buying of students when there is an attractive female present. Um, additionally, we might, um, take a look at this chart that's down here, um, of another example of naturalistic observation that occurs with technology today, with all the information that people can gather based on others, um, internet use. We can do a lot of naturalistic observations, um, by just pulling data from say, Twitter, um, pulling data from what people, how long people, um, until people swipe on a TikTok, um, looking at Instagram and the reactions people have, or how long they look at different, um, campaigns or ads that are present. There is a slew of free information that researchers have access to because of our increased use in telephones, in computers, and in social media.
So, for example, this is a snippet from our textbook where, um, a study was done where they looked at positive words in tweets and they looked at from midnight all the way like a 24-hour period, and then they broke it down by the seven different days of the week to see what, what time of day do we have the most positive tweets versus where do we have the most negative words in tweets. And so this will be something, for example, we can pull this chart out and say, "Okay, well, what time of the day do we have the most positive words? What time of the day do we typically have the most negative words? What day of the week do we have more positive words than negative words?" Again, I'm not able to draw any cause and effect here and not able to say that these, that Sunday is the reason we have positive words. All I can do is describe that, "Hey, here are the number of positive words that happen on a Sunday." Going back to that earlier slide on operational definitions, looking at this, something you should ask yourself is, "Well, how do they operationally define positive words in tweets?" What's a positive word to you? A positive word might be different than me, might be different than the researcher. So it's important that when we are reading a study like this, or any study, we are asking ourselves, "What are those positive words?" Do we agree with their operational definition? So we will practice looking at charts like this in class.
Third technique is a survey. Um, this, um, non-experimental method, um, you have done numerous times in school. Anytime you are given some sort of questionnaire where you have to give open-ended or close-ended answers. When we conduct a survey, it's really important that we choose wisely and get a representativeness, representative population or representative sample from our population. Um, so let's say we're doing a survey on high school students, um, at St. Charles North High School. Well, our population would be all high school students at St. Charles North High School. In order to get a good representative sample, I want to make sure I get individuals that represent each grade level, each gender, each, um, academic success, each, um, type of elective courses. You want to get a representative group of individuals to show a snippet of that population.
So when doing a survey, it's important to do an adequate random sample where you are giving every individual in the population an equal chance of being chosen to share their thoughts on a survey. So that would mean something as simple as putting every single name of the St. Charles North High School students into a random number generator and then spitting out, say, you know, 500 random numbers. Those numbers will be at random from the population, making sure that every single person in the population has an equal chance of being chosen. And by the end, it should be an accurate representation of the high school population. We want to avoid things like convenience samples where we just go and we do, um, third-period English classes. Well, that's a very easy way of getting a snippet of our population of high school students, but the problem is, is it's not a very representative sample. And we want to make sure that we are getting a representative sample when we're using surveys.
The second thing to keep in mind in surveys is the wording effect. And this is just to be a good consumer out there of research. When you're consuming research and you are, you know, reading what somebody is sharing with you that they discovered from their surveys, you want to be careful to look at, "Well, what was the question that they asked?" Um, oftentimes changing one word, making it just a little bit more biased, can impact the outcome of a research study. For example, saying the question, "Should your college allow speeches on campus that might incite violence?" versus, "Should your college forbid speeches on campus that might incite violence?" That difference between allow and forbid might sway the outcome. Or saying something like, "Do you believe in, um, a woman's right to choose?" or "Do you believe, um, that we, we shouldn't take the life of another child? Take the life of a child?" Depending on the words we use in our survey, the words we use in our questions is going to impact the outcome. It's going to impact how people respond to us. So we just want to make sure that we look at those words and make sure they're as neutral as possible when we're giving a survey, or they're as neutral as possible when we are being a consumer of what findings there were.
Um, finally, two more things with surveys. Um, one way to operationally define, remember, operational definitions need to be specific and measurable and allow for replication. One way to be specific and measurable is, let's say I am wanting to know, um, in a survey, your thoughts on, um, how effective your teacher has been in class. If I just say, "Tell me how effective," without using something like a Likert scale, it might make it difficult for me to compare and contrast results. Using something like a Likert scale, which you see here, is one to five, where I indicate one is strongly disagree all the way to five means strongly agree. I am giving a scale that everybody is using so that way when people are comparing their own thoughts to the Likert scale, they are applying it to the same exact scale. So I can compare the answers, males versus females, about that teacher, or, um, high achieving and low achieving, or A's, B's, C's, D's, and F students. I'm able to compare because I'm using that same Likert scale.
The last thing I need to be careful of in surveys is, um, what happens a lot of time is that, um, people want to look good. People want to, when they take a survey, they have this kind of social desirability. They want to look good. They want to, they might say things that make them look better and kind of have this, you know, social desirability, um, bias that might occur when they are, um, giving their responses in those. So we just want to make sure that we are taking survey evidence with a grain of salt, again, knowing it's a non-experimental method, so it cannot draw any causal statements. It can only merely describe to you human behavior.
And the last one is a correlational study. Which correlational studies are a part of naturalistic observations and survey methods. But correlational studies essentially just take two different things and see how they connect. So they see, um, how are height and shoe size connected? How are, um, political parties and snack choices connected? How are, um, gender and, uh, I don't know, clothing connected? They look at how two things relate to one another. So oftentimes, the naturalistic observation, we will pull a naturalistic observation to draw correlation. Again, I'm observing in the cafeteria how the, um, having an authority figure impacts snack choices. So I'm just going to merely look at, the more authority figures are, is there more or less snack buying happening? Um, or in a survey, I might again look at the correlation between, um, the what grade level you are and how much you liked a certain teacher. Again, I cannot draw any cause and effect and say because of this grade level, that's why you are liking this teacher or not, or because of this authority figure, that's why you're buying these snacks or not. I can merely just describe and say, "Hey, when there's more authority figures, snack buying decreases," or "Amongst juniors and seniors, this teacher is more well-liked."
Tomorrow, we are going, or our next video lecture is going to look at then that experimental method. We are going to look at how can we now draw cause and effect. So we've looked at how we, not non-experimental methods, and how can we describe behavior, but how can we explain why or where behavior comes from? And that's what, um, an experiment is going to do, and that's what we're going to focus on in our next video lecture.