Research 2 Exam 2 Study Guide

Question 1

What was the point of the video we watched?

Answer 1

The point of the video we watched was to show us an example of how a researcher had broken ethics in the way of fraud/fabrication of data. A student showed this to Lewandowski and it is a more interesting way to teach the topic of open science.

Question 2

What’s the difference between replication and reproducibility? Which is more likely to work? Why?

Answer 2

Replication means that we can use the data they provide, run what they ran, and receive the same results as the study. This is a check to make sure that they did their analyses correctly and did not complete an underhanded act of breaking ethics.

Question 3

Generally, what are questionable research practices?

Answer 3

Practices that are sketchy. They seem suspicious and the choice is not appropriate to a researcher.

Question 4

What is HARKing? b) Distinguish between testing hypotheses & exploratory findings.

Answer 4

Harking is hypothesizing after collecting data and analysis. In this questionable practice, they find data and then say oh yeah I expected this to happen. Liar, you just looked for a pattern and then claimed it all along. When our studies go through the IRB, we have to submit our hypothesis among other information in advance to any data collection. Testing your hypothesis means that you formulate your research question and then you set up your experiment, collect data, and then in your findings, determine if your hypothesis was supported or rejected. Exploratory findings are not created before the experiment, but done after you have the data. They are just some thing you wanted to look into or explore but not a hypothesis you test.

Question 5

What is p-Hacking? What are the 3 ways of doing that?

Answer 5

P-hacking is probability hacking, which means you affect your probability.
1.Cherry picking
2.Run your study until it works (e.g., analyze every 5 participants.
3.Play with variables to get them to work (e.g., selectively drop/include items)

Question 6

How often does falsifying data/fraud happen?

Answer 6

Not very often, it was reported 0.6 the BTS nerds who check the data found 1.7, although doubled, is still pretty low.

Question 7

What is preregistration? Why do we do that? Will this stop fraud/fabrication?

Answer 7

stating your hypothesis before conducting your study. Often in a public forum. We do it to show that out hypothesis was in fact created prior to the testing. It will prevent not stop fraud because there are multiple ways to commit fraud, this would only be stopping people from testing the study and then claiming that they knew it all along. It would not stop people from faking data.

Question 8

What are the three tenets of open science? Describe each. Which one helps with reproducibility? Which one helps with replication?

Answer 8

preregistration (keeps you honest), open data (every one can analyze your exact same data) reproducibility and open materials (replicability) can redo your study with their own participants.

Question 9

a) Give your own example of a double negative (including what it actually says). b) How does this apply to hypothesis testing?

Answer 9

a) I don’t disagree with you (I agree with you), you ain’t going nowhere(you’re going somewhere)
b) fail to reject the null
a) I am not never going to study. ( I am going to study)
b)In hypothesis testing, in order to reject the null we make a double negative. The null is that there is no difference, we show that there is not no difference (there is a difference) and we reject the null.

Question 10

What is the null hypothesis?

Answer 10

states there is no effect (change, difference) and the populations are the same The opposite of what we are hypothesizing. SImilar to a straw man argument we are setting it up to hopefully tear it down and support our hypothesis.

Question 11

Hypothesis test require that we start off assuming we’re wrong (with the null hypothesis). Give your own example of how this approach could be beneficial in another context.

Answer 11

This approach could be beneficial if we were in an ethical dilemma. If I thought it was possible my boss was committing fraud, but I did not have proof. It is better to assume I am wrong, and she is not committing fraud. THen I would talk to my boss and ask what they are doing. If it is fraud then I would report it, but if it isn’t then I end the process there. (If i assumed there was fraud when there was not (type 1 error) then they would have done all the extra processing and investigating to turn up with no fraud.

Question 12

a) List the 5 steps of hypothesis testing? b) In your words, describe what we’re generally doing in each. (we talked a bit about this in class, but also refer to the Table on pg 188)

Answer 12

a) First you have to label(establish) your population and your hypothesis. You need to know who you want to test/take a sample from and what you are looking for.
Then you need to build a comparison distribution
Third you establish the critical value cutoff
FOurth you determine the sample results
Finally you decide and interpret
b)state who you want to test/take sample from and what you are looking for (hypothesizing)
Then you need context so you have to see what or who you are comparing to. You should create a normal distribution.
Next, you establish the cutoff you must state what value or percentage you must reach in order for your result to count.
After that, you determine your sample results, conduct your study and receive the findings
Last, you compare the sample results to the cutoff and make a decision about the study.

Question 13

Distinguish between Type 1 and Type 2 error. Which is a false positive? Which is a false negative?

Answer 13

A type 1 error is when we claim something is true when it is false. (We say there is a ghost, when there is not ghost). This is a false positive because we think it is true but in actuality it is not. Fake true.
A type 2 error is when we claim something is false when it is true. We fail to reject the null when we should. (We say there is nothing there when there actually is a ghost). This is a false negative because we think it is false but in actuality it is not. Fake False.

Question 14

Interpreting p values. is this significant?
a) 5.43 4.47 N=1000

Answer 14

most definitely

Question 15

Interpreting p values. is this significant?
b) 5.43 4.27 N=63

Answer 15

probably not if 63 is small sample

Question 16

Interpreting p values. is this significant?
8.21 8.64 N=1,999,999

Answer 16

Most Definitely

Question 17

Interpreting p values. is this significant?
8.64 8.21 N=6

Answer 17

Definitely Not

Question 18

Interpreting p values. Is this significant?
5.55 5.73 N= 200

Answer 18

definitely not

Question 19

Interpreting p values. Is this significant?
6.78 9.91 N=200

Answer 19

most definitely

Question 20

Interpreting p values. effect size? significant? meaningful?
a) 5.43 4.27 p=.03

Answer 20

small effect, significant, not meaningful

Question 21

Interpreting p values. effect size? significant? meaningful?
5.43 4.27 p=.78

Answer 21

small effect, Not significant, not meaningful

Question 22

Interpreting p values. effect size? significant? meaningful?
8.21 8.64 p=.002

Answer 22

small effect size, significant, not meaningful

Question 23

Interpreting p values. effect size? significant? meaningful?
8.64, 8.21 p=.10

Answer 23

small effect size, Not significant, not meaningful

Question 24

Interpreting p values. effect size? significant? meaningful?
5.55 3.98 p=.000

Answer 24

Big effect size, impossible significance, meaningful?

Question 25

Interpreting p values. effect size? significant? meaningful? 6.78 6.03 p<.001

Answer 25

small effect size, significant, not meaningful

Question 26

Interpreting p values is this significant? d=.15 N=6000

Answer 26

most likely

Question 27

Interpreting p values is this significant? d=.85 N=58

Answer 27

Most likely

Question 28

Interpreting p values is this significant? d=.05 N=100

Answer 28

probably not

Question 29

Interpreting p values is this significant? d=.62 N=1547

Answer 29

most likely

Question 30

How does the mean of the distribution of means compare to the mean of the population of individuals? Why does this happen?

Answer 30

Equal to the mean of popul - all scores from og population. Taking samples plot for distribution, the mean is

Question 31

Explain in your own words how to create a Distribution of Means.

Answer 31

To create a distribution of means one would take the mean of every 5 people in a sample, over and over and then plot those means. Doesn’t always have to be five, taking samples mean of sample over and over.

Question 32

In your own words, what does a t-test for dependent means do/tell us? (i.e., looking at the formula, what is it doing? What does it compare?)

Answer 32

Compares within-subjects groups to see if there is a change from the first time the group is tested to the second time the same group is tested. For the t-test formula, we have the average difference subtract the comparison distribution to find difference, then divided by the average variability (with standard deviation)

Question 33

What types of variables to the t-test for Dependent means test analyze?

Answer 33

Nominal/categorical independent variables, interval ratio continuous dependent variables

Question 34

What research design does the t-test for Dependent means test analyze?

Answer 34

Strawman, like a z-score test? It is a t-test for dependent means THIS IS A WITHIN SUBJECTS DESIGN

Question 35

t-test for dependent means Give 2 examples of a research question that fits this design (one related to everyday life and one related to research)

Answer 35

Everyday life: Does coffee improve test scores? Research: Are people more willing to break the rules after doing so once?

Question 36

a) What is a difference score? b) How do you calculate it?

Answer 36

A difference score is the subtraction of the post to the pre or after to before. You calculate this by looking at the first time the group was tested and subtract that by the second time the same group was tested.

Question 37

a) What is the comparison distribution for the t-test for dependent means? b) What is the mean of this comparison distribution? c) Why?

Answer 37

a) The comparison distribution is a distribution of means of DIFFERENCE SCORES “sampling distribution” b) The mean is 0 c) because it always is, it represents the null hypothesis that there is no change from the first time the group was tested to the second time the same group was tested.

Question 38

Do you want to have a lot of degrees of freedom, or fewer degrees of freedom? Why?

Answer 38

Yes, you want to have a lot of degrees of freedom. The degrees of freedom are the number of the participants in the sample minus 1. We want our sample size to be large and more representative. A larger sample would mean a larger degree of freedom. Additionally, by having more df, our Distribution becomes MORE NORMAL.

Question 39

You run a study and find the mean difference score in your sample is 3.44. Is that a lot? How would you know?

Answer 39

Maybe? We do not yet know. The only way to know is with the context of the typical difference score. (If the typical is 3, than our sample is not that difference, if the typical is 1, than yeah our sample is pretty large.)

Question 40

You run a study and your sample's average difference score is 2.78. The comparison distribution has a mean of____and a standard error of 2.59. Is the difference in your sample big or small?

Answer 40

0; Small because we would be just slightly higher than one standard deviation from the mean. It is likely that our difference score is not big enough to pass the cutoff.

Question 41

a) If your study fails to reject the null hypothesis, what does that mean? b) What two things can you do to avoid this happening?

Answer 41

Our hypothesis was not supported by the results of the study. We did not reach the cutoff we needed for our test to be significant. increase sample size and increase effect size.

Question 42

When running a study, why do we include items that we don't plan on analyzing?

Answer 42

Some times we place distractor items within our study?

Question 43

When communicating results, what is the key with everyday language?

Answer 43

Write it as if your grandma or uncle could understand exactly what you are talking about without any statistics knowledge.

Question 44

t (53) = -2.27, p = .03, d = .42. Which number(s) is/are the degrees of freedom?

Answer 44

53

Question 45

t (53) = -2.27, p = .03, d = .42. Which number (s) is/are the significance level? Is it significant?

Answer 45

.03, yes

Question 46

t (53) = -2.27, p = .03, d = .42. Which number(s) is/are the effect size?

Answer 46

.42

Question 47

t (53) = -2.27, p = .03, d = .42. Which number(s) is/are the t score?

Answer 47

-2.27

Question 48

t (53) = -2.27, p = .03, d = .42. Which number(s) is/are the p level? Is it significant?

Answer 48

.03, yes

Question 49

t (53) = -2.27, p = .03, d = .42. How many tails is the test?

Answer 49

two

Question 50

t (53) = -2.27, p = .03, d = .42. In APA style, all statistical symbols are formatted in?

Answer 50

italics

Question 51

In your own words, what does a t--test for independent mean do/tell us? (i.e., looking at the formula, what is it doing?)

Answer 51

We are comparing between-subjects groups to find is they are from two different groups. Are there two kinds of people? The formula for the t-test is to subtract the mean of the first group to the mean of the second group divided by the standard difference (standard deviation)

Question 52

What types of variables does the t-test for independent means test analyze?

Answer 52

The IV are nominal/categorical and the DV are continuous (interval/ratio)

Question 53

What research design does this t-test for independent means test analyze?

Answer 53

This is a between-subjects research design

Question 54

Give 2 examples of a research question that fits this t-test for independent means design. (one related to everyday life and one related to research)

Answer 54

Everyday: Are college students who dress up for Halloween more fun than those who don’t dress up? Research: Are commuters more stressed than resident students?

Question 55

Please give your own examples of a) Pre-Post Design b) Two-group Design using the same topic/research question for both.

Answer 55

Here’ s an example… Topic: Happiness; Are people happier after they play with puppies? PrePost: Measure everyone’s happiness, let them play with puppies, measure again. Two-Group: Randomly assign some people to play with puppies while others don’t. Measure happiness of both groups) Topic: Creativity; are people more creative after they participate in a free-reign painting activity. PrePost: Measure everyone’s creativity, let them paint, measure again. Two-Group: Randomly assign some people to paint while others don’t. Measure creativity of both groups)

Question 56

If I give you the name of the comparison distribution, you should be able to tell me if it is for a t-test for dependent or independent means.

Answer 56

Distribution of differences between means = t test for independent means Distribution of difference scores = t test for dependent means

Question 57

a) What is the mean of the distribution of differences between means? b) What does this number signify?

Answer 57

0, this is the theoretical for no difference between populations, it represents the null hypothesis

Question 58

What is the t-test for independent means comparison distribution? How do we create it?

Answer 58

The mean of the distribution of differences between means is the comparison distribution for t test of independent means. We create it by estimating the population variance, then take the average, distribution of means, then the pooled estimate of population variance. We get the typical difference (sd) and use the mean of 0 to draw the distribution. (we would also show the t test cutoff)

Question 59

a) How do we learn the population's variance? b) What is the purpose of pooled estimate? c) How is the pooled estimate like the House of Representatives? d) Why do we do this (i.e., what does it accomplish)?

Answer 59

We learn the population’s variance by estimating, we use 2 between subjects samples that represent the population. b) Larger of the samples gets more influence, the purpose is to make this happen when necessary c) the pooled estimate is about the amount of influence each group gets. In the house of representatives bigger states with more people in them have more votes/say than small states with less people. d) we do this to make it fair, more people get more say than fewer people do.

Question 60

a) What is the formula for a t-test for Dependent means?

Answer 60

t= mean difference-comparison distribution/ standard difference (deviation)

Question 61

b) What is the formula for a t-test for Independent means?

Answer 61

t= mean of group 1 - mean of group 2/ standard difference (deviation)

Question 62

Remember your training from previous labs so that you can not only run t-tests but also run reliability/calculate means as needed.

Question 63

What Statistic Is This? Does watching a tutorial improve students' understanding of how to use EXCEL?

Answer 63

t test for Dependent means

Question 64

What Statistic Is This? A scale has 10 items that all allegedly measure charisma. Do they all measure the same thing?

Answer 64

cronbach's alpha reliability

Question 65

What Statistic Is This? How many students at Monmouth live on campus?

Answer 65

descriptives frequencies

Question 66

Who is cooler, people who have an iPhone or an Android?

Answer 66

t test for independent means

Question 67

How many patients at the local hospital have insurance?

Answer 67

descriptives frequencies

Question 68

Who has less anxiety, a group who did CBT therapy, or a group who did psychodynamic therapy?

Answer 68

t test for independent means

Question 69

Are 10 self-report items that measure narcissism reliable?

Answer 69

cronbach's alpha

Question 70

Do clients' depressive symptoms change from using relaxation techniques?

Answer 70

t test for Dependent means

Question 71

What is the mean of x-anxiety?

Answer 71

descriptives statistics (calculate mean)

Question 72

Does an athlete's mile time improve after drinking Gatorade?

Answer 72

t test for Dependent means

Question 73

At Monmouth, what's the average GPA of only Psychology majors?

Answer 73

select cases - descriptive statistics (average)

Question 74

Taylor sends 1989 texts a day. Is that in the top 5%?

Answer 74

z-score

Question 75

IV=Nominal variable with 2 levels-Within-Subject, DV=Continuous

Answer 75

t test for Dependent means

Question 76

IV=Nominal Variable with 2 levels-Between-Subjects, DV= Continuous

Answer 76

t test for independent means

Question 77

How is an ANOVA similar to a t-test for Independent Means? What is the key difference?

Answer 77

Similar because still comparing groups (are they the SAME population?) BUT the key difference is now we want to compare 3 or MORE groups.

Question 78

In your own words, what does an ANOVA do/tell us? What is the goal? What is the null?

Answer 78

Compare if there are differences in or between groups consisting of 3 or more groups. The goal is to find variance?

Question 79

a) What happens when you run multiple tests? b) What is an analysis plan? What does it help us do? c) How are exploratory analyses different from hypothesized ones?

Answer 79

You are being exploratory not testing your hypothesis. Analysis plan lists all the things we are going to run. It helps use visualize our steps and makes it easier to follow all steps. c) exploratory analyses are just looking for things, a hypothesized analysis is when you have previously stated that you are looking for something, and then you perform that operation (stat).

Question 80

In an ANOVA, where do our variance estimates come from? (i.e., what are the 2 sources?) b) When running a study, what do we hope is big? What do we hope is small?

Answer 80

Our variance estimates come from with a within-group or between-group design b) we hope the between-group population is big. We hope the within-group population is small

Question 81

What types of variables does this test ANOVA analyze?

Answer 81

IV = nominal/categorical DV = continuous (interval/ratio)

Question 82

What research design does this test ANOVA analyze?

Answer 82

This is a between subjects design? both?

Question 83

Give 2 examples of a research question that fits this ANOVA design. (one related to everyday life and one related to research)

Answer 83

this design. (one related to everyday life and one related to research) Everyday: Do people who like Sushi have less allergies than those who like pasta or chicken? Research: Does classical music, popular music, or rock music best help a plant grow?

Question 84

Where does within-group variation come from? (Give several sources with example for each)

Answer 84

Dumb luck, mood, experiences, personality DUmb luck, you could answer questions correct when you merely guessed or incorrect when you should have known them, people may have come in knowing how to dance, so when they are randomly placed they could be in the teaching dance group, but they already know it/find it easy, or they could be in the no treatment/control group and already knows how to dance scoring higher than most.

Question 85

The within-group estimate establishes the baseline. In your own words, what does that mean?

Answer 85

your own words, what does that mean? The baseline is the typical, the group come in with diversity and their own

Question 86

a) Give 3 group means with low variation. Give 3 group means with high variation. b) Which is more likely to represent the null? c) Which is more likely to depict samples that come from different populations?

Answer 86

Low variation= 3,4,5 High Variation, 1,4,7 b) the low variation is more likely to represent the null c) the high variation is more likely to depict samples that come from different populations.

Question 87

In your own words, please explain how we use sample means to estimate the between-groups variance.

Answer 87

We compare all sample means of the multiple groups together to estimate between-group variance.

Question 88

In your own words, what is sampling variation?

Answer 88

Always some due to change (when the null is true)

Question 89

a) Please draw 3 distributions that show between group estimation of population variance when the Null = True. b) Please draw another 3 distributions that show when the Null = Not True

Answer 89

a) like dog example b) like typical we hope it has more variance

Question 90

For the between-group estimate, the estimate can be different for two reasons. What are they?

Answer 90

Change factors The Effect/Treatment

Question 91

Be sure you're able to a) Generate Analysis Plans based on descriptions of data (like you did on Exam 1) b) Identify the key statistic based on a research question (like we've been practicing in class) c) Draw conclusions to a research question based on SPSS output (you've been doing this in lab)

Question 92

How are post-hocs and planned contrasts similar/different?

Answer 92

Post-hocs just look for anything but running everything planned contrasts look specifically for preplanned hypothesis, you want to know how a specific group compares to another or another two etc.

Question 93

a) Which is more like a fishing expedition, post hocs tests or contrasts? Why? b) Which do you use for exploratory work? c) Bonferroni and Tukey are both what? (hint: not delicious forms of pasta)

Answer 93

a) Post hocs, you just hope something bites b) post hocs are for exploratory work c) they are both post-hoc tests

Question 94

a) Using a picture (I'd give you something like what I did in class with the 3 cats and a dog), what would an ANOVA test? b) What would a t-test for independent means test? c) What would post-hocs test? d) What would contrast codes test the post-hocs couldn't test?

Question 95

Using the breakfast food study (Pop Tarts, Eggs, Coffee), show contrast codes for the following comparisons: a) Pop tarts vs coffee b) Eggs vs. Pop Tarts & Coffee) c) Pop Tarts vs. Eggs

Answer 95

POP Tarts Eggs Coffee a) 1, 0, -1 b) -1,2,-1 c) 1, -1, 0

Question 96

Based on this table, which groups are different? Pop Tarts 5.23(.18)ac Eggs 2.58 (.54) b Coffee 4.44 (.23)c N=93

Answer 96

Pop Tarts and Eggs are different because they have different subscripts, Eggs and Coffee too.

Question 97

Interpreting Statistics: F(2, 187) = 1.74, p = .048, η² = .09. Which number(s) is/are the degrees of freedom?

Answer 97

2,187

Question 98

Interpreting Statistics: F(2, 187) = 1.74, p = .048, η² = .09. Which number(s) is/are the significance level? Is it significant?

Answer 98

.048, yes

Question 99

Interpreting Statistics: F(2, 187) = 1.74, p = .048, η² = .09. Which number(s) is/are the effect size?

Answer 99

.09

Question 100

Interpreting Statistics: F(2, 187) = 1.74, p = .048, η² = .09. Which number(s) is/are the F score?

Answer 100

1.74

Question 101

Interpreting Statistics: F(2, 187) = 1.74, p = .048, η² = .09. Which number(s) is/are the p level? Is it significant?

Answer 101

.048

Question 102

How many tails is the test?

Answer 102

two

Question 103

In APA style, all statistical symbols are formatted in?

Answer 103

italics

Question 104

What statistic is this? Do exam grades improve from the first to second test?

Answer 104

t-test for dependent means

Question 105

What statistic is this? Are there awesomeness differences between students at Monmouth Rider, or Rutgers? If so, what are they?

Answer 105

ANOVA post-hoc

Question 106

Do psychology, anthropology, or sociology students send more texts? Do psych students send more than the others?

Answer 106

ANOVA and planned contrasts

Question 107

How many texts does a typical high school senior send?

Answer 107

descriptives - frequencies

Question 108

A new research student designs a study testing 23 different types of exercise on mood.

Answer 108

ANOVA post-hoc

Question 109

Based on midterm grades, do students do better in history or English?

Answer 109

descriptives - frequencies

Question 110

Among current students, are 4th year students less stressed compared to the 1st or 2nd year students?

Answer 110

ANOVA planned contrasts

Question 111

Explain the idea of signal to noise ratio as it applies to ANOVA.

Answer 111

signal is the between group and noise is the within group. There is a trend in the data but he have to separate the chance factors and variability that are between every individual.

Question 112

When the null is true, does that affect the between-group or within-group estimate? or both? why?

Answer 112

It affects the between-group because it means the treatment is 0.

Question 113

When the null is true, the between group and within group estimates of the population variance are essentially the same. why?

Answer 113

they are the same because the treatment is about 0. therefore only chance factors are shown in the between and within estimates leaving a total of 1.

Question 114

in your own words, what does the f ratio show?

Answer 114

how different your distribution is from the comparison

Question 115

a) draw and label the F distribution b) what does the shape tell you about F ratios? c) What might this suggest about life?

Answer 115

b) most fall around 1, few fall near 5 or higher c) most people are similar, no difference, only few are outside this and truly different.

Question 116

We discussed in class how thinking about the world in terms of within and between group variance is helpful. a) Give your own example of how you can apply these variance concepts to life. b) Which type do we pay too much attention to in life? Which type do we not pay enough attention to? c) Give your own example.

Answer 116

You want to push away the noise and only notice the treatment. I want to make sure that my treatment works and it isn’t the chance factors. WE think about differences between groups but have to consider similarities We often pay too much attention to the between group Between groups are fall season lovers and spring, winter, summer season lovers. Within groups is all people who have different personality, mood, experience, and chance luck.

Question 117

Your thesis on how different breakfast foods influence stat exam scores compares Pop Tarts, Organic/Farm Raised Eggs, and Black Coffee on test scores. Your ANOVA test comes back significant. You do a dance of joy (obviously). a) What does the significant result tell you? b) What does it not tell you? (related to your study)

Answer 117

It tells you there is a difference between pop Tarts, Organic/Farm Raised Eggs, and Black Coffee on test scores. YOu do not know where the difference is. WHat specifically makes the impact on test scores. Is it pop tarts? Eggs? Coffee? Is one better than the others?