Okay, so let's look at the last ten questions. Question 21, what ethical aspect is definitely violated? Well, the right answer here of course is respect. Students weren't even aware that they were taking part in a research study. So they did not autonomously and voluntarily decide to take part in the study. And of course they were not well informed. Question 22. Suppose the lecturer does not find the expected results and decides not to reject the hypothesis yet, but to reject the assumption of valid and reliable measurement instead. By whom was this approach critically discussed? Well the right answer here is Quine of the Duhem Quine thesis. Who said it is always possible to keep a hypothesis alive by rejecting a background or supporting assumption or hypothesis. This is in reply to Popper who favored falsification. In which of course for the reason I just said doesn't work according to Quine. Kuhn also talked about hypotheses that should have actually been rejected, but stayed around too long. But this is always in the context of certain paradigms and paradigm shifts. Question 23, the choice to measure engagement with a questionnaire is associated with the deduction phase in the empirical cycle. In the induction phase, you transform observations into hypotheses, a general rule. In the deduction phase, you use this general rule in combination with decisions about operationalization, manipulation, a research design to deduce specific outcomes, specific observations that form your predictions. In the testing phase you actually collect the data and see whether your predictions are confirmed. So the right answer here was deduction. 24, if an item from the engagement questionnaire is considered to be applicable by highly engaged but also highly unengaged students, this item is not monotone. Because a higher score is not unambiguously associated with more engagement. That terms polytomous and dichotomous are associated with a number of response options and are not relevant here. Question 25, suppose the research hypothesis is not confirmed by the results and the lecturer decides not to publish the research for that reason. This is an example of. Well, you could say publication bias. It is that, but in this course in the videos I made a further specification, the file-drawer problem, for failures to publish reports based on the researchers decision. Now I realize and I saw that in the forum there's been a big discussion about this. And the, the use of this term is, is not unambiguous. Many people use it in different ways. But in this course, it's presented in the way I just described, and also the question performs very nicely in terms of psychometric qualities. So, I decided to leave this one in, and count as the only right answer the file-drawer problem. Question 26. If the engagement questionnaire is very unreliable, then the next time any particular students complete the questionnaire, what will happen to the new score? Well consider what happens if you stand on a very unreliable weight scale a couple of times. Well you would expect large variations every time. This is basically the definition of unreliable. So the right answer is by a large degree. Question 27 what kind of variable is type of student? It's not monotone. We discussed the term monotone item, but not monotone variable here. This is not relevant. It's not an experimental variable, because we didn't manipulate whether somebody is a premaster or bachelor student. Individual-differences variable is the right answer. It's a characteristic that participants bring with them into the research study. What is the measurement level of the variable educational degree? Well, it could be nominal because we specified different classes or categories, you can see that hear. Secondary you get education, bachelor, or master's degree. And of course these build on each other, so there is actually an order to these degrees. [SOUND] So the right answer is ordinal, not interval. You can't say anything about the differences between these levels. Okay, so moving on to question 29. Now this was a big mess of a question because you actually didn't have the relevant background information. The on campus students were privy to extra material explaining this in more detail. So we decided to award you a point for whatever option you choose. But to give you an idea of what my reason was here, if you look at these statistics you can see that the minimum value was one and the maximum value was three, and that missing values were specified. If you go back to the description of the variable sex in the research description you. You can see that there were two categories, male and female. No third category. Don't want to say or anything like that. So there should be only two values. And the most obvious coding would be one and two given these values here. So three is strange, is weird. And because its adjacent to the number two on your keyboard its very likely a data entry error. It could maybe be a missing value misspecification. But it would be very, well basically stupid to use the value three as a missing value. For example if people indicated they were both male and female and you wanted to indicate that you can't use these these data. They should be considered missing. That's a really bad idea. Why not use nine or zero or minus one, or something that's can't be mistaken or, or, or isn't easily mistyped and therefore erroneously considered a missing value. So, the actual right answer was, a data entry error. A coding error is very unlikely here, because why would you recode that? Unfortunately people were thrown off by these data, the mean and standard deviation which is weird information to provide for a categorical variable. But my idea was that it was a really quick way to see that the, the, the average was kind of high but, but you should have been able to to answer this question without this information. It was made up. Not really relevant. Okay. The last question, number 30. The benefit of a syntax-file, amongst other things, is documentation of. Well of course it's documentation of the statistical operation performed on the data. You don't want to include the actual materials, the actual questionnaires and stuff like that in your syntax file. So the right answer here was the first option and that's it. Okay, now let's check my results. Yes I'm sure. Yay! 30 out of 30. Okay, so there you go. I hope this clears up any last questions that you might've had about the final exam questions.
