Fundamentals of Design & Statistics Exam - May 2021 - Psychology

This document presents the solutions to the Fundamentals of Design & Statistics (FDS) exam for a Psychology module. The exam covers various aspects of statistical analysis and research design, including creating datasets with skewed distributions, interpreting statistical outputs (correlations, ANOVA), and evaluating research methodologies. The solutions address questions on experimental design, including factorial designs, independent and dependent variables, and the interpretation of statistical findings. Ethical considerations, such as informed consent and the use of deception in research, are also addressed. The solutions also cover the challenges of order effects in repeated measures designs and the selection of appropriate statistical tests (parametric vs. non-parametric) based on data characteristics. The assignment further assesses the ability to identify and explain issues within proposed research projects and the interpretation of statistical results in APA format. The solutions provided demonstrate an understanding of key statistical concepts and their application in psychological research.

TRIMESTER TWO EXAMINATION – MAY 2021
MAIN
COLLEGE: Social Sciences and Humanities
DEPARTMENT: Psychology, Sociology and Politics
MODULE TITLE: Fundamentals on Design & Statistics
SUBMISSION DEADLINE: 18th May 2021 at 12:00 GMT
SUGGESTED DURATION: 48 hours
___________________________________________________________________
EXAMINATION CONDUCT:
1. The University Academic Conduct Regulation outlines the behavioural
expectations of candidates completing any examination.
2. Students are responsible for ensuring that they know how to submit their exam
script, when the deadline is and that they submit the script in enough time before
the deadline expires. It is anticipated that Blackboard will be slower around
submission times.
3. It is a fundamental principle that students are assessed fairly and equitably. The
University Academic Conduct Regulation defines unfair behaviour relating to an
examination to be 'cheating'. The University will investigate and may sanction any
acts or behaviours which breach the Code of Academic Conduct.
4. In view of the University Covid-19 response and the exceptional actions necessary
to conduct off campus exams, all students are reminded that this is an individual
task and that students who contact or collude with other students to complete their
exam may be subject to sanction later.
5. The criteria to Request to Repeat an Assessment Attempt (RRAA) has been
extended to include Covid-19 related issues - this includes illness, self-isolation, IT
issues due to remote working, accessibility of learning and research materials,
unexpected childcare.
INSTRUCTIONS TO CANDIDATES:
1. Whilst the release period for this exam is a 48-hour period, this exam should take
you no longer than the above suggested duration to complete. Please be sensible
about the amount of time you spend completing and submitting your work and
make allowance for technical issues prior to the deadline.
2. Candidates can use their own calculator.
THIS PAPER CONTAINS 13 PAGES INCLUDING THIS SHEET Page 1 of 13

MODULE CODE: 77-701752
3. This is an OPEN BOOK examination. You may find that your FDS workbook will
be of use in this exam.
4. Academic support will be available in three 1-hour slots between 12-11pm on 18th
May, 12-1pm on 19th May and 11am-12pm on 20th May.
5. It is possible that you may encounter technical issues during the exam; if you have
any difficulty with IT you should consult the below student guidance document on
My Hallam which contains useful information on hints and tips, contact numbers
and links to support:
6. Any changes or clarification to the exam paper will be communicated via the
module Blackboard site announcements. It is recommended that students monitor
Blackboard announcements prior to submission of their final script but particularly
in the first hour after release of the exam paper.
___________________________________________________________________
STATIONERY REQUIREMENTS PER CANDIDATE:
None – use this document to return your typed answers
___________________________________________________________________
Page 2 of 13

MODULE CODE: 77-701752
1. In jamovi, create a dataset that exhibits a significantly skewed distribution. Provide evidence of
this by obtaining the following measures and pasting the outputs below, and explaining how
each metric demonstrates this:
 Skewness statistic (and standard error)
 Shapiro-Wilk analysis
 Histogram or density plot
 Boxplot
Page 3 of 13

MODULE CODE: 77-701752
2. Please read through the following statistical output and answer the questions at the end:
A researcher is interested in the relationship between the degree of personal ambition and work ethic
amongst city traders on the London Stock market. He wishes to control for Machiavellianism (a
personality trait) as a possible confounding variable, and duly collects the data. Assuming the data met
the appropriate assumptions for parametric analysis, he conducted a set of zero-order correlations
followed by a partial correlation (see tables below).
(Zero-order) Correlation
Personal
Ambition
Work
Ethic
Machiavellianis
m
Personal
Ambition
Pearson's
r —
p-value —
Work Ethic Pearson's
r 0.71 —
p-value < .001 —
Machiavellianis
m
Pearson's
r 0.39 0.45 —
p-value < .001 < .001 —
Partial (First order) Correlation
Personal
Ambition Work Ethic
Personal
Ambition
Pearson's
r —
p-value —
Work Ethic Pearson's
r 0.65 —
p-value < .001 —
Note. controlling for 'Machiavellianism'
 Citing the full statistical evidence (APA convention) to substantiate your explanation,
summarise what has been found:
It can be found that the P value is .065 which is more than P= 0.05. thus, null hypothesis is rejected. It
means that there is relationship between the degree of personal ambition and work ethic amongst city
traders on the London Stock market. Here, the trader personal ambition does not depend on work ethic
(Babbie, Wagner & Zaino, 2018). Their ambition is different from work ethics. Moreover, high goal are
set by them which they want to achieve in future.
Page 4 of 13

MODULE CODE: 77-701752
3. Please read through the following statistical output and answer the questions at the end:
A College tutor investigated the effect that watching television might have on the motivation
levels of her students. She engineered a study in which one group could only watch ‘Daytime TV’
programmes (such as This Morning), another group watched a series of TED Talks, and a 3rd group
had their televisions present but switched off. The analysis is below:
 What type of design is this?
It is a observational research design.
 How many participants were included in the analysis?
The number of participants were 87.
 Why was a Welch’s ANOVA used? Support your answer citing relevant evidence.
It compares two means to find out if they are equal or not. So, here it was used to compare means of
daytime tv and ted talks (George, & Mallery, 2019). With that it was easy to find out and measure
means. Also, it determine mean value of two or more groups easily. So, in above case there are tww
groups that is why welch’s anova is used.
 Based on the output, write a paragraph to interpret the study citing statistical evidence (APA
format) with a summary conclusion.
Page 5 of 13

MODULE CODE: 77-701752
It can be interpreted that P value is 0.024 that is less than P= 0.05. this means that there is no the
effect that watching television on motivation levels of students (Stehlik-Barry, & Babinec, 2017). Even if
children watch motivational programs they do not get motivated. It might be due to that they just watch it
for fun. Moreover, there are many other factors of motivation for children. Also, motivation depends on
children mood as well while watching motivational programs.
.
4. You are a supervisor listening to your project student presenting their intended dissertation
project.
They state that they wish to design “a study to prove that mindfulness is better than drinking
alcohol at reducing stress”. In order to do this, they plan to get one group of students to engage in
an hour of mindfulness each morning – whilst the second group will not do this. Instead, this
second group will be required to keep a log of how many units of alcohol they consume each day.
Both groups will do this over a fortnight period, and at the end of this will need to answer a
questionnaire that measures stress levels.
As the supervisor, point out and explain 5 issues with this as described to you:
1. There is no hypothesis made.
2. No time mentioned that how many hours are there.
3. No criteria for measuring alcohol consumption
4. No criteria for measuring stress levels
5. There is no characteristics of group
Page 6 of 13

MODULE CODE: 77-701752
5. The graph below charts how many participants in a within-subjects design (two conditions)
would be needed to achieve a power level of 0.8 (80%) using a standard alpha cut-off of p
≤ .05 and a moderate effect size (d=0.5).
Power Curve by N
 Explain what would happen to the projected sample size (and why) if instead a stricter cut-off
value of p ≤.005) was anticipated at the planning stage
The projected sample size might have been small as the value of P is correlated with sample
size.
Page 7 of 13

MODULE CODE: 77-701752
6. Order effects are particularly problematic for within-subjects/repeated measures designs.
Explain why this is and the basis behind a method of your own choosing to minimise the
potential confounding effects of this.
This is because order effect are caused by exposing subject to multiple treatment. Moreover,
there can be change in measure design by which it order effect is a problem in it.
Thus, the statistical method can be used to minimise potential confounding effect. This method
is used because it is more effective than non statistical method. The outcomes obtained from
using statistical method are also accurate and it minimise confounding effect.

Page 8 of 13

MODULE CODE: 77-701752
7. You have been asked to design a research study investigating the efficacy of prescribing
mescaline (a known psychedelic drug) on reducing the intensity of intrusive thoughts in a study
lasting 2 months. You decide to implement a factorial design to test this in healthy controls vs.
patients with psychoses. You decide to compare mescaline against placebo and against
clozapine (a commonly used anti-psychotic). You decide to collect scores on the severity of
intrusive thoughts both pre-, during- and post-intervention.
 What type of a factorial design is this? Use the numbering system to illustrate this.
It will be a 3*2*2 factorial design.
 How many Independent Variables (IVs/factors) are in this design?
There are 2 independent variables in this that are mescaline and clozapine.
 How many levels have you created in each IV/factor?
There are 3 levels in it that is pre, during and post
 How many conditions are in this study in total?
There are 2 conditions that is healthy control and patient with psychoses.
Page 9 of 13

MODULE CODE: 77-701752
8. As part of any study, those individuals recruited to take part have to first demonstrate that they
can give their ‘informed consent’. Explain then how it is possible to make use of deception in
research.
It shows in study that participant with their own wish and consent had participated in study. Also,
consent from state that data is valid and no false info is included in research. Moreover, it might
happen that consent form is wrongly taken from participants and then incorrect data is entered into
it. In consent form info written is fake (Allen, Bennett, & Heritage, 2018).thus. the scholar may use
fake info written in consent form and then provide invalid outcomes from it. Besides, some false info
can be added into research by misleading the participants.
Page 10 of 13

MODULE CODE: 77-701752
9. You and your research partner are reviewing data collected from 12 clinical patients and
whilst comparing the measures of central tendency you both note that the mean is some
distance away from the median.
 Explain what the most likely reason is for this:
Here, the reason can be that values are not distributed normally. Also, it shows that median is
better representation in this case. Besides, it might be due to fact that data when set from
low to high is not proper. There might be some value or data added into it as well.
 Due to this same reason, you decide that the data should be analysed with a non-parametric
inferential statistic. Your partner looks aghast and wants to push ahead with the parametric
equivalent. Explain below two (statistical) reasons that you could use to try and persuade
them.
 The reason are as below:
 First is inferential statistic will provide accurate and relevant outcomes as compare to
parametric one. It uses measurements from the sample of subjects in the experiment to
compare the treatment groups and make generalizations about the larger population of
subjects.
 Another is even if parametric validity is not met the non parametric statistic can be applied. It
will allow to take findings from sample group and then generalise in proper way.
Page 11 of 13

MODULE CODE: 77-701752
 “There is no such thing as a mixed-methods approach to research. It is one or the other”.
 Explain what the mixed methods approach is and illustrate one way in which this could
actually be implemented:
Mixed method approach refers to research that advance systematic integration of qualitative
and quantitative data in study. Here, the scholar uses both methods to collect data. The
qualitative is used to gather secondary data from various sources and quantitative can be
used to collect data through questionnaire. In that only stats and facts are gathered. Thus,
these both mixed methods approach is used to study. It can be actually implemented by
either using qualitative or quantitative research approach.
Page 12 of 13