 This is part three of lecture two in this final part of the lecture We're going to be talking about different research designs that you can do when you conduct your work So there's basically two Research designs that I want to talk to you about today That's first of all the correlational methods and secondly It's the experimental method and it's very important for you to know the difference between the two so pay attention So when let me first start by talking about the correlational methods If we use this method what we can then do is examine the variables the the naturally occurring relationship between variables, so we are not influencing one variable. We're not trying to to Impact a certain variable and no we just look what happens if we if we see how variables naturally Relate to each other. So let me clarify this by giving you an example of Correlational research. Let's imagine you have the following research question. Do extroverts have more friends than introverts? So you probably are familiar with the trait extroversion So some people are more extroverted and some people are more introverted and you might be thinking to yourself I think people with who are more extroverted have an easier time connecting to others And are more assertive and maybe they also have an easier time making new friends So my prediction would be that extroverts have more friends than introverts So if this is your research question, it would be a wonderful question to address by using a correlational design So what you can then do is use a survey so a questionnaire in which you basically just simply ask participants About the number of friends that they have and you have them fill out an extroversion questionnaire And then the outcome could be that extroversion is positively Correlated with the number of friends and then this would be in line with your hypotheses So a positive correlation a positive relationship between two factors between two variables That would show that these these factors indeed relate to each other So if you use a correlational design, which you then find is An outcome that you can plot in a graph here You see several possible outcomes of a correlational research This is of course not real data. This is just an illustration So what you first see is a perfect positive correlation. This is something that you never find an actual research So if you find this then you're probably related to the Dirk Stadel because this just doesn't show up in real life So just for you to know this is not what we actually find But this is just for illustrative purposes here you would find a relationship of one a perfect positive correlation So this means that if you know one factor So for example, if you know the extra version score of people then you can perfectly predict how many friends they have So the only thing you need to know is their extra version level and you know that the more extroverted They are the more friends they have it's a positive correlation and a perfect correlation of one again This never happens. So on the very opposite side, you see a perfect negative correlation and here this is a correlation of minus one and This is also again a perfect correlation in the sense that you only need to know one factor And then you can predict the outcome on the other factor So in this case with our example, this would be the more extroverted you are the less friends you have So a negative relationship between the two variables of minus one So the the plot you see in the middle. That's zero correlation. There's no relationship between the two factors This is actually something that we come across as scientists a lot Oftentimes we have hypothesis. We have ideas and they turn out to be you know Untrue we cannot find support for them in the data set So this is something that I came across a lot in my research career, which is perfectly fine We have ideas we test them and sometimes they turn out to be incorrect So here there's zero correlation That means that the level of extroversion is completely unrelated to the number of friends that you have So you can find support or you can reject your hypothesis based on the Correlation coefficient that is always somewhere between one and minus one and if it's close to zero Then you can reject your hypothesis Okay When you do correlational research, there's a lot of advantages. You can just study two factors. You don't have to intervene. You don't have to Have very complicated Paradigms you don't even have to go to the lab often you can just ask questions and see how Variables relate to each other, but there's a very big downside of using a correlational design And that is that you only see that there's a relationship Two factors have something to do with each other But you do not know whether one is causing the other you do not know if one factor is Basically the cause and the other is the consequence or the other way around So you just know that there's a relationship between the two just like you probably would find a relationship between Ice cream sales and sunburn. So how much sunburn there's been sold in the stores. There's a relationship there I think we can all understand why right? It's not that buying ice cream always sort of spurts is Create you or makes you go also to buy sunscreen No, there's something else going on and the underlying factor is of course the weather So the cause is actually the hot weather and that's causing people to both Buy ice cream and both get a sunburn or buy sunscreen Okay, so we see a correlation between two factors, but that's not saying that they have a causal relationship So that's very important to understand if we do Correlational research we cannot talk about cause and effect we can only only talk about a relationship between two factors and We as social psychologists We do not really like this because what we really want to do eventually is predicting behavior We want to understand why people do what they do and we want to sort of get a grip on Predicting people's final behavior. So what we love to do is conduct experimental research So what is experimental research? This is research in which we create groups create groups of individuals And all the participants in our research are randomly assigned to the groups Can be two groups or three groups sometimes more of it than it gets a bit messy So oftentimes in a research that uses experiments, we have two or three Conditions, that's what we refer to two or two three groups that are created by the researcher and Then we measure something so we create groups we influence one factor and then we measure another factor So let's again give an example for further clarification Let's imagine that you have a research question that is dust temperature So how cold or hot participants are those that affect how helpful they are So let's imagine that that's your research question That will be a beautiful question to address in an experiment in this experiment You can divide participants between groups and you can ask participants to either take place in a cool room or a warm room If you do so it is very important that you use random Assignments and with random assignments I mean that you do not come up with a reason why some people sit in the cool room and others sit in a warm room For example your research would not work if you say, okay All the men go to the cold room and all the women go to the warm room Because what what you're doing then is not random assignments. You're using assignment based on another factor So if you find a difference between conditions Then you never know whether it is because people were in a cold or a warm room or because all the men were in the cold room And all the women were in a warm room So you should never do that you should use random assignments but over the groups and then You can test helpfulness you can for example do so in a clever way in which a researcher enters either the cold room Or the warm room drops a box of pants The participants is sitting there and you just gonna check how many pens the participant picks up So is this participant actually going to help the researcher when the researcher drops the pens? Will the subject help to clean up and is there a difference between helpfulness behavior in the cold room versus the warm room? Okay, this is a fake study just for you to remember so this is not an actual question But this could be a research design that you could actually do The advantage of using this method if you use random assignments is that you can actually make claims about causality cause and consequence because you're actually Influencing one factor in this case temperature You're sending people either to a cold room or a warm room and you're testing a difference in helpfulness So here it is really clear that you have two variables One is the cause and one is the consequence and we have specific names for them in our experimental designs So we have one independent variable and this is the variable that is affected by the researcher that the you the researcher Is is sort of manipulating and the researcher assumes that this factor is the cause So in our little example, this would be temperature Temperature is the independent variable and there's another variable that we're interested in and that is the dependent variable This is the variable that is measured and the researcher assumed that this is the consequence So this dependent variable, it's already in the name right dependent is dependent on the first variable on the independent variable In our example, this would be helpfulness. How helpful are participants? Okay, let's look at the fake results of this fake study Here you see what could happen in your results Let's imagine that this would be what your results look like on the y-axis You see the number of pens the participants pick up and on the x-axis You see whether people were in either in the warm room or in the cold room and you see that here the bars are different Right in the warm room participants pick up about seven or eight pens and in the cold room about three or four pens So you see that there's a difference Then what researchers always do is check the probability level the P level and you can do so with by using various statistical techniques But this P level that's vital information because this P level informs us About the chances that the results of an experiment are actually due to chance and Not because the independent variable differs between conditions. So this P value We like this to be as small as possible ideally Smaller than one point oh five and in this in our experiment our fake experiment if P value is one point oh one And what this means is that this? In this example the probability that the results that we get is actually due to chance That is one in one hundred So there's a one and one hundred percent chance that the results that we find here is due to chance So that's actually that's good. That's a we call this a significant results It's statistically significant results. So the P value is something that you should keep in mind So let's imagine we find this result and you think okay cool. Well, this is solves right? We know that there's a difference. It's a statistically significant difference. So question solves the only problem with the design that we use in this study is That we only have a warm room and a cold room So we do not exactly know which of the two rooms is actually causing the effect Is it that's people in a warm room become more helpful or is it that people in a cold room become less helpful? And that's a difference So we are both basically manipulating the factors in both groups and that's not ideal This is actually sort of and not an ideal a situation. So let's improve this experiment Let's create three groups one group with a warm room The second group in the cold room and the third group in a room in which temperature is unaffected This so so there's no this just a normal temperature for participants to be in not affected by external factors By adding this condition. This is called a control condition We can actually see whether people become more helpful in the warm room or less helpful in a cold room So let's imagine we add this condition and this is what the results look like What is this saying to us? What how how can we interpret this? Results so here you see that people in the control condition pick up about the same number of pens as participants in the cold condition And what we can then conclude is that being in a warm room makes you actually more likely to help Being in a cold room doesn't do much, but it's being in a warm room that really increases helpfulness behavior If the results looks like this the conclusion is different So let's imagine the control group the people in the control group behave about the same as people in the warm Space then the conclusion is no, it's not that being in a warm room makes you more helpful No, it's being in a cold room which makes you less likely to help and that's a vitally different conclusion Okay, so this is all that I wanted to share with you today. Thanks for watching this lecture