PSY202 AUG 2020 Open Book Final Exam Answer Sheet
Some reminders:
I have left comment boxes assigned to each question for marking purposes. If they bother you, you can hide them by going to Review > Tracking > Change the dropdown menu from “All Markup” to “Original”. Otherwise, please do not remove the comment boxes.
Name :
Student ID :
Section A
Q. No. | Answer |
Q1. | The addition of barriers does not significantly increase the variance explained in hand-washing behavior beyond that of the other predictors because they account for 2.4% increase in the variation (R^{2} = 0.024) |
Q2. | Without controlling for perceived behavioral control, subjective norms, and barriers, regression analysis shows that attitudes and handwashing behavior do not have statistically significant relationship, as indicated by zero order correlation coefficient (r = 0.351, p = 0.324) |
Q3. | The regression model does not meet the assumption of homoscedasticity because the scatter plot shows unequal distribution of standardized residuals of hand-washing behaviors |
Q4. | The regression equation: Hand-washing behavior = 0.766 (PB) + 0.059 (ATT) + 0.08 (SN) – 20.593 |
Q5. | Two suggestions to enhance hand-washing behavior among healthcare workers i) Improving perceived behavior control (p = 0.000) ii) Reducing barriers to handwashing (p = 0.000) |
Q6. | R-square change (0.024) is the same as squared semi-partial correlation (0.155^{2}) because they both indicate the degree of variation in the handwashing behavior explained by barriers when other predictors are controlled in the regression model. |
Section B
Q. No. | Answer |
Q1. | Game explains 31.8% of the variation in the level of mental dexterity (partial eta square = 0.318). |
Q2. | The mean difference (0.085) in mental dexterity between the Wizard’s Chess group and the Control group has the confidence interval with the lower limit of 0.0403 and the upper limit of 0.1297. By using a significance level of 99%, the confidence interval of the mean difference would increase (CI > 0.0403-0.1297). |
Q3. | Matching allows pairing of samples in different treatment groups to those in the control group to minimize effects of confounders. In this case, the Professor should match individuals with the same intelligent quotients to minimize its confounding effects. Randomization entails selection of samples by chance to eliminate the selection bias. The Professor should employ randomization in selecting individuals with matched intelligent quotient and putting them on the treatment and control groups. |
Q4. | The graph supports the hypothesis 2a that participants who play Wizard’s Chess will have a bigger improvement in mental dexterity (before and after playing the game) compared to the other two groups because the slope of the line is the steepest and the post-test mean is the highest (M = 0.9418). Moreover, the graph supports the hypothesis 2b that participants who play Quidditch will have a bigger improvement in mental dexterity (before and after playing the game) compared to the control group since the post-test mean is relatively higher (M = 0.8190) although the sloppiness is the same. |
Q5. | Two potential explanations for parallel means in the graph i) The participants had different levels of intelligent quotients ii) There might have been confounding variables on the control group that had the same effect as the game of Quidditch on mental dexterity |
Q6. | Professor McGonagall should recommend Professor Longbottom to employ the Bonferroni test as an appropriate post hoc test because it decreases the probability of false positives. In essence, this test decreases the probability of rejecting the null hypothesis. |
Section C
Q. No. | Answer |
Q1. | Since the coefficients of total effects (β = -12.421, p = 0.000), direct effects (β = -10.074, p = 0.019), and (β = -2.857) indirect effects are statistically significant, mediation is feasible. Partial mediation exists in the model because the demand and the burnout accounts for the variation in the level of productivity. The mediation index is 0.189, which accounts for 18.9% of the variation in job productivity. |
Q2. | Lord Pusheen can test mediation using hierarchical regression analysis to test effects of job demand (path a) on burnout, then burnout on job productivity, and direct effects of job demand on the productivity (path c’). The changes in the significance of the regression coefficients when burnout is added would indicate the existence of mediation. |
Q3. | Mediator influences the effect on a predictor on a dependent variable, while confounding variables affects both the dependent and independent variables in a given study. |
Q4. | |
[SEC A: Q1]
[SEC A: Q2]
[SEC A: Q3]
[SEC A: Q4]
[SEC A: Q5]
[SEC A: Q6]
[SEC B: Q1]
[SEC B: Q2]
[SEC B: Q3]
[SEC B: Q4]
[SEC B: Q5]
[SEC B: Q6]
[SEC C: Q1]
[SEC C: Q2]
[SEC C: Q3]
[SEC C: Q4]
Department of Psychology
August 2020 Semester
Online Open Book Final Examination
Subject Code: PSY 202
Subject Name: Advanced Quantitative Methods II
Instructor(s): Mr. Timothy Liew
Release of Open Book Final Exam questions (Date & Time) via LMS: 3
rd December 2020, 3:00pm
Deadline to submit answer scripts (Date and Time) into Turnitin: 4
th December 2020, 3:00pm
Instructions
1. This is considered an online open-book final examination, where students are allowed to consult
notes and textbooks.
2. We recommend that you take the exam in a quiet space, free from interruptions and disturbances,
in the comfort of your home, with a reliable internet connection.
3. APA referencing and reference lists are not required, much like how they are not required during
conventional final exams.
4. However, this does not mean that students can copy and paste directly from other sources, or
discuss answers with / copy from each other.
5. All responses have to be in the students’ own words, and if content is taken from another source, it
must be paraphrased.
6. Students found with high similarities to each other or other sources via Turnitin will receive
penalties and be called for academic misconduct.
7. A forum will be open in LMS for the original duration of the final exam (e.g. if the exam was
originally from 9-11am, the forum will be open for those 2 hours). This is so students can clarify any
uncertainties with the exam content or exam instructions (if any).
a. Students who have any questions should post their questions on this forum, and the lecturers
will respond within the duration stipulated above.
b. Please do not email lecturers with your questions, but post them on the forum.
c. Please also check the forum for questions that have already been answered, prior to posting
potentially similar questions.
d. Please refrain from asking content-specific questions that can reveal your answers in the LMS
forum.
e. Questions posted on the forum after this duration (or emails sent after this duration) will not be
entertained.
Submission
1. Here are the main steps to submit your answers:
▪ Step 1: Download the prepared answer sheet and write your answers there. Remember to
also fill in your name and student ID within the document file.
▪ Step 2: Once you’re done, rename the file to “YOUR STUDENT ID + SUBJECT CODE”. For
example, B1234567 – PSY202
▪ Step 3: Submit your renamed file to Turnitin (25603902, eggsbenedict). Please submit your
final exam answers there. Please do not change the answer sheet into a pdf file; just submit
it as a Word document.
2. No submissions will be accepted once the deadline of 24 hours (from the time of the release of the
exams questions) is over.
Page 2
3. Please keep a copy of your Turnitin submission acknowledgment receipt (sent to your email; the
acknowledgment receipt is also downloadable on Turnitin). Your submission will only be considered
as accepted once Turnitin has reverted with this acknowledgment receipt.
4. If your upload was not successful and you do not receive an acknowledgement of submission from
Turnitin, please try uploading your assessment again.
5. If you still run into any issues with submission, please email your lecturers immediately to notify
them of the issues you’ve experienced, prior to the deadline. Please include screenshots / videos
(with timestamps clearly visible) of the issue you are facing.
6. Please do not delete / remove yourself from any of your subject Turnitin pages.
Additional Instructions
1. This paper consists of 3 sections:
▪ Section A: (6 questions, 18%)
▪ Section B: (6 questions, 18%)
▪ Section C: (4 questions, 14%)
2. This online open book final examination constitutes 50% of the final grade, with a total of 50 marks.
3. If a question asks you to draw a diagram (spoiler alert: only one question asks that), feel free to use
whatever method suits you (e.g., using power point to draw the diagram and then saving it as
image, drawing it out manually and then taking a photo of it, dazzling me with your photoshop
skills, etc.)
4. Limit your answers to four pages (I don’t foresee you needing more than four pages to answer the
questions). Please ensure you follow the submission instructions in the previous section.
5. Good luck!
Page 3
PSY202 Online Open Book Final Examination
Section A (18%)
Professor Dovin would like to improve hand hygiene among healthcare workers. To help frame his
study, Professor Dovin relies on two models as the foundation of his theoretical framework, the primary
one of which is the Theory of Planned Behaviour (TPB). TPB is a model that describes how one’s belief
and behaviour are linked and is frequently referred to in the context of health psychology. Specifically,
TPB is constituted by three main constructs that predict an individual’s intention to perform a certain
behaviour:
1) Attitudes – The degree to which a person likes or dislikes the behaviour,
2) Subjective Norms – The social pressure a person feels to perform the specific behaviour, and
3) Perceived Behavioural Control – The degree to which a person feels that they have sufficient physical
and or cognitive resources to perform the behaviour
In addition, Professor Dovin also suspects that barriers to handwashing might also explain poor handwashing adherence, as the Health Belief Model postulates that perceived barriers might prevent
engagement in the target health behaviour. Barriers to handwashing may include time constraints,
inaccessibility or insufficient number of sinks, and harsh cleaning agents.
Armed with this knowledge in hand, Professor Dovin recruits 500 participants which includes doctors,
nurses, and various other healthcare workers. He disseminates the TPB Construct Scale, which is made
up of three subscales that measure attitude, subjective norm, and perceived behavioural control. Total
scores are obtained from each subscale, where higher total score indicates more favourable attitudes
towards hand-washing, higher social pressure to wash hands, and higher perceived behavioural control
to perform hand-washing, respectively. Furthermore, Professor Dovin also administers the Perceived
Barriers Subscale, where higher total scores on the subscale indicates higher perceived barriers (i.e., the
individual perceives that there are a lot more barriers to hand-washing). Finally, Professor Dovin
measures hand-washing behaviour via the Wash-Yo-Hands-Bruh Scale, where higher total scores
indicate more frequent and effective hand-washing.
After acquiring his data, Professor Dovin runs a hierarchical multiple regression. He enters the three
constructs of TPB into the first model, and then inserts perceived barriers into the second model.
Page 4
Note on labelling
PBC : Perceived Behavioural Control
SN : Subjective Norms
ATT : Attitudes
Barriers : Perceived Barriers
Descriptive Statistics
Mean Std. Deviation N
Hand Washing Behaviour 23.1706 15.74500 381
PBC 54.7767 12.84142 381
ATT 54.8605 13.75843 381
SN 18.0773 12.40091 381
Barriers 62.1063 14.90114 381
Model Summaryc
Model R
R
Square
Adjusted R
Square
Std. Error of
the Estimate
Change Statistics
DurbinWatson
R Square
Change
F
Change df1 df2
Sig. F
Change
1 .628a
.395 .390 12.29600 .395 82.025 3 377 .000
2 .647b
.419 .413 12.06625 .024 15.494 1 376 .000 1.393
a. Predictors: (Constant), SN, ATT, PBC
b. Predictors: (Constant), SN, ATT, PBC, Barriers
c. Dependent Variable: Hand Washing Behaviour
ANOVAa
Model Sum of Squares df Mean Square F Sig.
1 Regression 37204.598 3 12401.533 82.025 .000b
Residual 56999.281 377 151.192
Total 94203.879 380
2 Regression 39460.387 4 9865.097 67.757 .000c
Residual 54743.492 376 145.594
Total 94203.879 380
a. Dependent Variable: Hand Washing Behaviour
b. Predictors: (Constant), SN, ATT, PBC
c. Predictors: (Constant), SN, ATT, PBC, Barriers
Page 5
Coefficientsa
Model
Unstandardized
Coefficients
Standardized
Coefficients
t Sig.
95.0% Confidence
Interval for B Correlations
B Std. Error Beta
Lower
Bound
Upper
Bound
Zeroorder Partial Part
1 (Constant) -20.593 3.091 -6.663 .000 -26.671 -14.516
PBC .766 .063 .625 12.181 .000 .643 .890 .623 .531 .488
ATT .059 .053 .052 1.106 .270 -.046 .164 .351 .057 .044
SN .080 .057 .063 1.397 .163 -.032 .192 -.203 .072 .056
2 (Constant) -8.556 4.307 -1.986 .048 -17.025 -.086
PBC .750 .062 .612 12.130 .000 .629 .872 .623 .530 .477
ATT .052 .052 .045 .989 .324 -.051 .155 .351 .051 .039
SN .104 .056 .082 1.844 .066 -.007 .215 -.203 .095 .072
Barriers -.166 .042 -.158 -3.936 .000 -.250 -.083 -.242 -.199 -.155
a. Dependent Variable: Hand Washing Behaviour
Page 6
Page 7
1. Does the addition of barriers significantly increase the variance explained in hand-washing behaviour
beyond that of the other predictors? Include any relevant statistical evidence in your answer. (2%)
2. Is there a statistically significant relationship between attitudes and handwashing behaviour without
controlling for perceived behavioural control, subjective norms, and barriers? Explain your answer
and include any relevant statistical evidence if necessary. (2%)
3. Based on the output, would you say the assumption of homoscedasticity has been assumed? Explain
your answer and explicitly state which part of the output are you basing your answer on (4%)
4. Please provide the regression equation for Model 1. (3%)
5. Given the results from the statistical output, provide TWO suggestions on how to enhance handwashing behaviour among healthcare workers. (4%)
6. You have learned previously about the semipartial correlation, sr. You have also learned recently
that you can square it to obtain the squared semipartial correlation, sr2
. Now, go ahead and obtain
the sr2 for the perceived barrier predictor (round up to three decimal place), and then compare this
statistic with the R2
-change statistic in Model 2. Notice how they are the same value? This is not a
coincidence; I would like you to unleash your critical thinking and explain why does the sr2
for
perceived barrier and the R2
-change statistic in Model 2 are equivalent. (3%)
Page 8
Section B (18%)
Being a secret master of games, Professor McGonagall is interested in seeing how different types of
games may improve a person’s mental dexterity (the capacity of the mind to think creatively and quickly
understand something). She looks through the list of wizard games and decides to settle three groups:
❖ Wizard’s Chess
▪ Like regular chess but the pieces move on their own woooooo.
❖ Quidditch
▪ You know Quidditch right? It’s like when two teams fight soccer on flying broomsticks (look
it up if you must).
❖ Control Group
▪ These participants will dutifully play no games
Professor McGonagall proceeds to recruit her participants from the different Houses in Hogwarts. She
promptly divides them into Groups 1, 2 and 3 without much fuss. Before having them play one of the
three wizard games, the good professor measures their mental dexterity. How does she do so? Quite
simple actually! The professor instructs the participants to perform the Daedalus’ Investigative Map
(DIM) exercise, which involve solving a complex puzzle that changes every time you attempt it
(magic!). The less time you take to solve the puzzle, the better your mental dexterity (for example, smart
bean Hermione Granger managed to solve the puzzle under a minute).
After measuring their pretest mental dexterity, Professor McGonagall instructs her participant to engage
in the different type of games for one hour per day for a total of three weeks (i.e., Group 1 participants
play Wizard Chess, Group 2 plays Quidditch, and Group 3 plays no games). After three weeks, Professor
McGonagall gather her participants and again instructs them to perform the Daedalus’ Investigative Map
(DIM) exercise to measure their posttest mental dexterity after playing the games.
Right before the experiment began, the professor noted that participants in Group 1 and 2 were
particularly excited, and seem to already expect their mental dexterity to improve after three weeks of
playing their respective games.
Once that’s done, she hires a muggle statistician to analyse her data in order to answer her hypotheses:
H1: There would be a main effect of time, whereby participant’s mental dexterity will be higher after
playing the game for two weeks compared to before playing the game
H2a: Participants who play Wizard’s Chess will have a bigger improvement in mental dexterity (before
and after playing the game) compared to the other two groups.
H2b: Participants who play Quidditch will have a bigger improvement in mental dexterity (before and
after playing the game) compared to the control group.
Page 9
Descriptive Statistics
Game Mean Std. Deviation N
Pretest Control Group .6739 .09840 18
Quidditch .6834 .09979 20
Wizard Chess .7103 .06991 29
Total .6925 .08771 67
Posttest Control Group .8082 .10299 18
Quidditch .8190 .08518 20
Wizard Chess .9418 .06880 29
Total .8692 .10451 67
Tests of Within-Subjects Effects
Measure: Mental_Dexterity
Source
Type III Sum of
Squares df Mean Square F Sig.
Partial Eta
Squared
Time Sphericity Assumed .898 1 .898 117.340 .000 .647
Greenhouse-Geisser .898 1.000 .898 117.340 .000 .647
Huynh-Feldt .898 1.000 .898 117.340 .000 .647
Lower-bound .898 1.000 .898 117.340 .000 .647
Time * Game Sphericity Assumed .077 2 .038 5.006 .010 .135
Greenhouse-Geisser .077 2.000 .038 5.006 .010 .135
Huynh-Feldt .077 2.000 .038 5.006 .010 .135
Lower-bound .077 2.000 .038 5.006 .010 .135
Error(Time) Sphericity Assumed .490 64 .008
Greenhouse-Geisser .490 64.000 .008
Huynh-Feldt .490 64.000 .008
Lower-bound .490 64.000 .008
Tests of Between-Subjects Effects
Measure: Mental_Dexterity
Transformed Variable: Average
Source
Type III Sum
of Squares df Mean Square F Sig.
Partial Eta
Squared
Noncent.
Parameter
Observed
Powera
Intercept 76.757 1 76.757 10872.362 .000 .994 10872.362 1.000
Game .211 2 .105 14.922 .000 .318 29.844 .999
Error .452 64 .007
a. Computed using alpha = .05
Page 10
Multiple Comparisons
Measure: Mental_Dexterity
Scheffe
(I) Game (J) Game
Mean Difference
(I-J) Std. Error Sig.
95% Confidence Interval
Lower Bound Upper Bound
Control Group Quidditch -.0101 .01930 .872 -.0585 .0382
Wizard Chess -.0850*
.01783 .000 -.1297 -.0403
Quidditch Control Group .0101 .01930 .872 -.0382 .0585
Wizard Chess -.0749*
.01727 .000 -.1181 -.0316
Wizard Chess Control Group .0850*
.01783 .000 .0403 .1297
Quidditch .0749*
.01727 .000 .0316 .1181
Based on observed means.
The error term is Mean Square(Error) = .004.
*. The mean difference is significant at the .05 level.
*Note: First line from the top represents Wizard Chess, followed by Quidditch, and Control Group
Page 11
1. Which statistic would tell us the proportion of the variance in mental dexterity that is explained by
type of game? Identify and report that statistic. (2%)
2. Report the confidence interval for the mean difference in mental dexterity between the Wizard’s
Chess group and the Control group. Next, explain and elaborate on what would happen to the
confidence interval if Professor McGonagall had specified a 99% confidence interval instead. (3%).
3. Prior to her study, Professor McGonagall was concerned about the influence of individual
differences; in particular, she was worried that her results would be confounded by participants’ IQ.
Please define and describe how the Professor could use matching and random assignment as control
measures, and how would this help reduce the confounding influence of IQ. (3%)
4. Interpret the interaction graph and relate it to H2a and H2b. Your answer must contain specific
descriptions of the graph (e.g., steepness/direction of the slopes). (4%)
5. Based on the interaction graph and the scenario, provide two potential explanations for the pattern of
results in relation to H2b. Be sure to adequately elaborate your answer. (4%)
6. Professor Longbottom, a colleague of Professor McGonagall, has approached her for advice. The
school of Hogwarts has instructed Professor Longbottom to investigate the effect of the type of
potion on a person’s physical health. He has been asked to compare 6 different potions, and because
this is related to physical health, the school explicitly requested Professor Longbottom to prioritise
avoiding false positives over false negatives.
Given the descriptions above, which post hoc test should Professor McGonagall recommend
Professor Longbottom? Explain and justify your answer (2%)
Page 12
Section C (14%)
Lord Pusheen believes that while productivity and getting things done is important, they also firmly
believe that this should not come at the cost of one’s mental health. Acting on that conviction, they
decide to conduct a study to examine the relationship between job demands (i.e., the demands placed
upon an individual in an occupational context) and productivity (i.e., how effective and efficient a person
is at completing goals and tasks). Taking a cue from the Job Demands-Resource Model, Lord Pusheen
hypothesises that the proposed relationship can be explained by the experience of burnout. After
establishing the theoretical framework, Lord Pusheen recruits participants from the realm of Pusheenia
and measures their levels of job demands, productivity, and burnout.
**************************************************************************
Model : 4
Y : Productivity
X : Job Demands
M : Burnout
Sample
Size: 180
************** TOTAL, DIRECT, AND INDIRECT EFFECTS OF X ON Y **************
Total effect of X on Y
Effect se t p LLCI ULCI c_ps c_cs
-12.421 .473 -26.287 .000 -13.353 -11.489 .462 .892
Direct effect of X on Y
Effect se t p LLCI ULCI c’_ps c’_cs
-10.074 .819 -12.295 .019 -11.680 -8.449 .375 .723
Indirect effect(s) of X on Y:
Effect BootSE BootLLCI BootULCI
Burnout -2.857 .677 -3.623 -1.027
Partially standardized indirect effect(s) of X on Y:
Effect BootSE BootLLCI BootULCI
Burnout -.088 .025 -.137 -.038
Completely standardized indirect effect(s) of X on Y:
Effect BootSE BootLLCI BootULCI
Burnout -.169 .049 -.265 -.073
*********************** ANALYSIS NOTES AND ERRORS ************************
Level of confidence for all confidence intervals in output:
95.0000
Number of bootstrap samples for percentile bootstrap confidence intervals:
5000
—— END MATRIX —–
Page 13
1. From the output below, can we infer that mediation has occurred? If so, what is the type of
mediation? Explain your answer and include the supporting statistical evidences, including the index
of mediation. (4%)
2. The Gods of Statistics have decided to play a prank on Lord Pusheen, and have forbidden the use of
PROCESS 3.4 to test for mediation (oh no!) Based on what you’ve learned, describe how can Lord
Pusheen test for path a, path c, and path c’ without the usage of PROCESS 3.4. Make sure you
make reference to Lord Pusheen’s variables in your answer. (4%)
3. Describe the difference between a mediator and a confounding variable. (2%)
4. Lord Pusheen wants to better understand the relationship between burnout and productivity. After
discussing with Sir Pip, Lord Pusheen hypothesizes that the relationship between burnout and
productivity may differ depending on resilience, where the relationship between burnout and
productivity may even approach non-significance with higher levels of resilience.
Based on the description above, please draw out the statistical diagram for Lord Pusheen’s new
model (4%)