Simple Experiments and Causal Claims, Exams of Nursing

Download Simple Experiments and Causal Claims and more Exams Nursing in PDF only on Docsity!

TEST BANK

With 100% Approved Answers

3_23_Simple_Experime

nts

How so? This is the class for Monday, April 23rd. And today we're going to start talking about simple experiments. So We're going to be talking. I'm going to give an example of a simple experiment. We're going to talk about experimental variables and why experiments support causal claims. And then in the next class we're going to get into independent groups designs. And then the following class, we're going to discuss within group designs in interrogating causal claims with the 4 abilities. So here you're learning objectives, please read them over and this is what I would like you to leave the class, knowing as well as what you should know for the test. So this is an example of a simple experiment. example of a simple experiment. Researchers were interested in knowing what's a better method for for learning. Is it taking notes by hand or is it typing on a laptop? a laptop? You know, oftentimes we in classrooms, you see both methods unless you know the teachers, they know laptops and there are some, there's some empirical support for no laptops as well, but anyway, so laptops, you know, there are some benefits you get to write everything that the professor is saying, however, when you're writing, you're not really thinking about the content. You're just kind of writing it down, But for writing, maybe you know by by longhand maybe you are not really writing every word that the instructor saying, but you're writing summaries and you're drawing pictures. And so therefore you're having a deeper processing of the materials. So there are some positives and negatives to both concepts, but there was a study that wanted to examine this. So they compared the effectiveness of taking notes In class and but they didn't actually test us in class. They tested and in a lab. So they recruited $67.00 students to come into the lab and half the time when they walked into the lab, there were laptops there. And half the time when they walked into the lab, it was notebooks and pens. So they were randomly assigned to a condition and the 2 conditions were either a laptop or pen and paper. And then everybody despite their conditions, they watched a ted talk on an interesting topic. So Ted Talks are if you don't know, it's like a video that an expert is coming in and he's or she is talking to an audience about About some sort of interesting research that's going on in the field right now. And they talk for about 5 to 20 minutes, talks are really interesting. If you get the chance to watch them, I would highly recommend that they speak in plain language. So it helps you don't have to know like big conceptual ideas in order to understand it. But anyway, so they watched a ted talk and then they were told to take notes using their assigned method. So the people that were in the pen and paper method, they wrote long hand notes and then they completed a distractor task for about 30 minutes. So this was that way they weren't thinking about the content of the video destructor tasks. Some common ones are like completing a word search, doing a puzzle, doing anagrams, things like that, just something that would get your mind off of whatever's happening. And then they were tested on what they learned. So they completed a test that had both factual and conceptual. I know, so factual is like, you know, understanding definitions and things like

And so those are the 2 types of conditions that there were. There were 2 conditions or 2 levels within the independent variable of types of taking. And then the deep end of variable is the measured outcome variable. So this was a learning and you can have, you can have more than one d.v.d. If so in in the study that we just talked about, 's. there were 2 D.V.D. There was the factual and conceptual learning, but you have less control over the dependent variable. You have full control over the independent variable, but then the whole point of the study is to see what happens when you manipulate the independent variable, how that affects the variables. So you really have no control over that variable., So the i.v. the independent variable comes 1st and then the d.v. Comes after. I mean, you can think about that as like I view. So I v. In then the d.v.d. Right.

. So I have a kind of looks like the number one. So I see this comes. 1st looks like one comes 1st. So that's a way to remember it. And the deep end and variable is deep pendant on the independent variable. So that's kind of a way to think about that too. Just little things to help you remember which one is, which So control variables are anywhere you pull that and experimenter holds constant and we need to control 3rd variables. So we do this in order to control for independent In order to have high internal ability. So I hope you can hear me drinking anyway. So so far we've only talked about statistical control. So this is like the multiple regressions where you're adding multiple variables in and controlling for them and seeing if the relationship still exists when controlling for those variables. So we have control variables in experiments as well. And this is because we need to make sure that we're varying only one thing at a time and had a limit alternative explanations for for our findings. And like I said in establishing his internal validity. So in the note study, they had a control variable, where people in both conditions and both groups watched the same videos and answered the same questions. So they held that they held that constant. The the content in the videos weren't different because if the contents in the videos were different, well then maybe it's the video that's Affecting things rather than the note taking somebody holding the video and by holding the questions constant, they don't That they can make sure that what they're finding is truly due to the notes. So these are just, this is like a quick overview of what we're going to be talking about. So experiments establish covariance. So again, covariance is saying that the causal variable or the cause variable is

related to the effect variable. So in this case with the note study, the type of note is related to the amount learned. Experiments also establish temporal precedence. So were were asking, does the cause a variable come before the the dependent variable, the thing that we're measuring. So did the type of note taking come before the learning or before the arm? The well, yeah, I guess the learning would be. And then finally, well designed experiments that's important because you can have an experiment that kind of sucks and you don't establish internal validity, but do while design while design Spearman's establish internal validity. So are there any alternative explanations for results? If there aren't, then we have established internal validity. So let's get into experiments establishing covariance. So We can't know so sorry. Let me let me back up a little bit. So covariance is indicated by a difference in group means. So what I mean by that is we had 2 conditions, right? We had the long hand and the laptop group and we found that the Long Hand group was higher in learning. So there is covariance there because there's a difference between the groups. If there were no differences like how we found no differences, I guess we didn't find about how they found that there were no differences in the factual group. There's no covariance there. But within the conceptual questions there were differences. So there is covariance that's been established And Independent variables into the question compared to what. So we need comparison groups. So in the in the note taking a study, they used laptops. I'm just going to repeat this a 1000000 times. Apparently they use laptops compared to Compared to hand written notes, but In other studies, we need, we need comparison groups as well Because you can't know if If you have covariance, unless you compare it to something else, you can't find differences between groups without comparing a group to something else. So for instance, if you think that a large pasta bowl is making you eat way too much, pasta, just because of the size of it, you can't know if that's happening in less. You eat pasta in a smaller bowl because then you're comparing it to something else and you're seeing the difference between how much pasta you eat. So it's hard to determine without some sort of comparison group. Another example is there was an article on Psychology Today that a researcher had put out about how they had analyzed pictures of dogs being hugged on the Internet. And they found that 82 percent of dogs showed signs of distress when they were being hugged. But there were a couple of problems with this 1st off. They, they didn't. That's not a peer reviewed study that they were talking about. Also, they didn't compare

All right, so now we're going to talk about Different Different things that threaten internal validity. So these are all ways that we can get alternative explanations for things. And we need to rule out those 3rd variables because like I said, they threaten internal validity. So in the notes study, a confound, so confound, is it something that confuses the cause? So a compound in the note study would be Where some participants watched a really boring video and took notes on that were while other participants watched a really interesting video and took notes on that. So just because The quality of the video that would that would affect the amount that people retain the information right compared to the notes. So Our Ivy, the type of know it's is confounded or confused with material quality. We don't know what's causing what's affecting the test scores, is it because like our people are some people in the laptop group getting lower scores because they watched a boring video so they didn't care to to encode the information in their brain. We can't really tell that if we have some confounding variable, so here Notes with interesting material. They score higher on the test and lap top notes with morning material. They score lower on the test. But again, we can't tell if that's due to if that's due to the type of notes, or if that's due to the type of material. So a design confound is, a 2nd variable that various systematically along with the i.v., the dependent variable, and produces an alternative explanation. So In the, in the notes study, if the questions that were harder in the laptop group than they were in the notes group, then we wouldn't know if the difference was due to conceptual performance. Or if the differences in conceptual performance were due to The type of notes, or if it was due to the the questions like how hard the questions were. So again, we can't really isolate the cause if we have these confound variables in there because it's just an alternative explanation. Right? It threatens the internal validity and we can't support the causal claim. And systematic very variability is the problem. So systematic variability, this is where This is where something systematically is different between the groups. So for instance, if the longhand group, the one writing by longhand, if they all watched very interesting videos, whereas the the laptop group watched really boring videos, then that would be, that would be systematic variability. Unsystematic variability is not a confound where as systematic variability. So this unsystematic variability is where there is random, It's like random differences, random variability within the the groups, and it affects both groups. So it's not a confound. So this might be that, you know maybe there are the people are watching everyone's watching the same video, but people in both groups just

aren't really interested in the content. Like maybe you, the video was about age and Egypt. Right. And some people and both groups were alike injured. Egypt's really boring. I don't care about it. And they just that They process the information differently. But if it's happening in both groups, that's not a problem. And one issue with a systematic variability is just that it makes it can make differences between groups, kind of hard to recognize. They can make it difficult to detect differences. But The unsystematic variability is important to have compared to systematic variability. And that's why random assignment is so important. So selection of facts is another type of Variable that can or design that can affect Internal validity. And so this is where participants in one level are in one level of the independent. Variable are systematically different from another. So this is where participants essentially they choose their own condition And it can make it very. It can make it hard because participants aren't randomly assigned. So therefore, there is systematic differences or systematic variability within the groups. So an example of this is that there was an intensive therapy done for kids with autism and some got a new treatment or some continued with their own old treatment. And then they compared the 2 groups, but they weren't randomly assigned because of the distance that some people had to go to travel for for the intensive therapy as well as how long the intensive therapy was. It was like a full time job to be a part of this intensive therapy. It was like 40 hours a week and that just wasn't possible for some people. So they chose their own group in this case. And as you can see the those in the treatment as usual, they scored lower in improvement than those in the intensive treatment condition. So the intensive treatment conditions showed more improvement, but because there were selection effects, it's really hard to know. Well, is it due to the treatment or is it due to maybe the parents, you know, the parents that are willing to be there are 40 hours a week at work with their kids that way. Maybe they're just a little bit more eager, more motivated to to help their children as awful as that sounds. But so this is a threat to internal validity. We can't determine the reason for the results, is it because of the treatment or is it because of the type of parent? And we can avoid selection of facts with random assignment. So random assignment is where I, You randomly, assign participants to each level. And this make sure that each participant has an equal chance of being in either group And there's no systematic differences between the groups. Typically, there was a study that was done that found that random assignment creates similar groups, 98 percent of the time. So there's really no differences between the groups and then we just have, we have void having systematic effects or systematic variation or variability, which is not good. Right. Though, that's some sort of confounding variable. So you can see those here. So we have a population.