Okay, so let's look at question number 11 through 20. Question number 11, the lecturer is definitely not a, well definitely not a rationalist. Because a rationalist believes that knowledge is best acquired through reasoning. And the lecturer holds that knowledge about students' engagement is best acquired through measurement or observation. So not a rationalist. She's also not an idealist, because idealists believe that reality exists in our minds dependent on human thought and the lecturer holds that engagement really exists independent of human thought. So the lecturer is not a rationalist and not an idealist. So she's not both, a rationalist and an idealist. Neither you can also say. Question number 12, suppose the expected effect occurs only for premaster students. The lecturer reports the result of this group of participants only. What do we call this? Well, this is not harking, this is not changing your hypothesis after the fact. It's not checking for preliminary data and then deciding to stop for example. It's cherry picking. It's choosing the data that looks pretty, that agrees with your hypothesis and presenting only that, those results. Question number 13. Suppose the lecturer decides to add sex as an independent variable. What does this mean for our design? Well, that means it's now factorial, because we have not only type of feedback as an independent variable but also sex, so we have two. It's not a within or within between because repeated exposure of participants to different levels of these independent variable is not involved. So factorial is the right answer here. Question number 14. Possible threats to internal validity, against which the research design does not protect, are. Well, mortality because drop, mortality or dropout is very hard to mitigate. The best you can do is document dropout very carefully. Take procedural measures to make sure that you motivate participants as much as you can to stay in the study. Of course within ethical boundaries. But that's, that's basically all you can do. Also history yes this is also something in that, that is very hard to mitigate. To to make sure that no unforeseen effect occurs. Well there, there, there isn't any design that, that helps with that. So in this case the right answer was both of these threats are not mitigated through the design that was used in this step. Question number 15. Okay, suppose that in population of women, the variation in engagement scores is greater. So there are more high and low scores for the women as compared to the men. Now if we take a sample of the same size for men and a sample for women, what will this do with the precision of the estimate of mean engagement for women as compared to men? Well the precision of course will be smaller, will be less because these women are all over the place. They have really high and really low scores. So there's a good chance that a, any particular sample will show a less accurate estimation of the true mean engagement. So precision is smaller. Number 16. The lecturer could have approached the research question qualitatively by. Well, by using an open interview. A structured interview is basically a questionnaire. It has predefined set response options. It might be more complicated. It might be face-to-face, conducted by an interviewer. But the response options are predefined by the researcher. This is not the case in an open interview, where it's up to the participant where the interview goes, what subject would be discussed and what the final answers are. This corresponds, or characterizes a qualitative approach. So the right answer here is an open interview. Question number 17. The lecturer could have performed a randomization check by checking if. Well, checking if the comparative feedback was accurate, that's, that's a manipulation check, not a randomization check. That's an alternative to the manipulation check that we saw earlier. Engagement was the same between groups at the end of the course. Well, this is not the right option because this option describes the final effect that we're interested in. The effect of independent variable feedback on dependent variable engagement. Is it the same? Well, we hope not. We hope that the comparative feedback improved the engagement in the experimental group. So we are left with this option, and this is the right one because it, it al, it's also about the dependent variable, but at the beginning of the course when no manipulation has taken place yet. You hope that through randomization. On average engagement will be the same between groups at the start of the study but we could have measured this at the beginning of the study by doing a pretest to make sure that there are no pre-existing differences that could explain a difference at the end of the study and would thereby form an alternative explanation. So this is the right answer. Question 18. Suppose the exam grade does not give a good indication of knowledge of statistics. This means it's not valid, it has, has low construct validity. Well then of course, the answer to this question is the problem is for construct validity of the exam. Of course the research hypothesis is not really concerned with the exam grade. It's about the causal relation of type of feedback on engagement. Now, if exam grade wasn't measured accurately, this doesn't really impact the internal validity. Whether the causal relationship between independent and dependent is actually true. And it doesn't impact whether this relation is generalizable. [SOUND] Okay. Question number 19. For fundamental research, external validity is. Well, the answer I was looking for was less important than internal validity. Because in fundamental research, we usually or usually the, the, the general aim is to show that there is a certain causal relation. We usually investigate these relations in convenient samples. In highly artificial settings. So external validity is usually somewhat lower. This is not really a problem be, because usually, initially in fundamental research the aim is to show this causal relation. So I, I messed up here. I should've included initially in the question. Because I didn't I can imagine that you might of also chosen as important as internal validity because in the end, if our relation turns out not to have external validity, well, then it means that it's not reproducible in other people or in other settings or other times. Which means there's something missing from our hypothesis. There's another factor at play that explains the differences between these situations or groups. So that's why we decided to give you a, a point, no matter, no matter what what option you chose, because the question was flawed. Okay, question number 20. The most obvious threat to external validity in this study is, well of course it's selection. because we have a highly selective group of people that are not representative of all students everywhere. But you could have reasoned out this question another way. There's no mention of an unforeseen event so history is, is less likely than selection. And so is reactivity, reacting to the fact that you're being studied as a participant because the students weren't aware that they were taking part in a study. So, the right answer here is selection.