Assessment Brief and Marking Criteria 2023-24

Module title Applied Statistics and Data Visualisation

CRN G340 M0001 / CRN41140

Level 7

Assessment title
Statistical Analysis and Interactive
Dashboard Design
Weighting within This assessment is worth 100% of the overall module mark.
module

Submission deadline 06/12/2023 at 4:00 pm.

date and time

Module Leader/Assessment set by

Kaveh Kiani, email: [email protected]
Nathan Topping, email: [email protected]

How to submit
Your assessment should be submitted through blackboard and should be separated into
two formats. First, a single pdf report (between 5k-7k words) and second, a zipped file
containing the R codes and Power Bi dashboard. Please check that the report file and zip
file are:

1. Your report has been named as “your name.pdf” like ”John Smith.pdf”.
2. Check the zip file is valid and openable.
3. The zip file should contain material that are clearly labelled and fully working
versions of R codes and Power Bi dashboard should be included with a clearly
written description of each application and its use in a “Read Me.txt” file. Your
dashboard should be shared as a .pbix file.

1
Assessment task details and instructions:

Your task is to demonstrate your newly developed knowledge and understanding of data
handling, validation, statistical analysis, and visualisation by exploring and presenting data
from an extensive and complex data set.

There are two sources of data for this assignment:

1. World Development Indicators (WDI)
2. UNdat
World Development Indicators (WDI) and UNdata are the main World Bank collection of
indicators and United Nations Data Bank, compiled from officially recognised international
sources. Both sources include national, regional, and global estimates. Both sources include
numerous indicators for countries around the world. Also, there are yearly slices of data
for the countries that could be construed as time series.

The dataset for this assignment can be accessed from one of these two sources of your
choice:

https://databank.worldbank.org/source/world-development-indicators

http://data.un.org/Explorer.aspx

Once you have followed the above links, you can download the dataset by selecting the
countries and variables and years you want to work with.

This assessment requires a comprehensive statistical analysis, and a working dashboard

prototype to be presented. Your statistical analysis and interactive dashboard design should
be fully justified and explained in your report. The assessment also requires you to
demonstrate the justified use of techniques for data preparation, validation, analysis and/or
modelling and prediction; appropriately referencing research into dashboard composition,
layout, function, and form. Justification of the approaches taken for statistical analysis and
visualisation is expected and outputs should be provided. Your reasoned thinking, research,
and critical evaluation of both the problem resolution and your solution also form a
substantive part of this work.

2
Assessment Tasks

The assessment has 2 tasks, Statistical analysis and interactive dashboard building. You
should complete these 2 tasks and also do other necessary steps that have been mentioned
in the “Report Writing Frame” document.

Imagine you are working as a Data Scientist at a Non-Government Organisation involved in

social and economic development globally. Part of your role is to use data to do statistical
analysis to predict some events or test some hypothesis also to communicate your
interesting findings to a wider public.

At your first task at the organisation, you are given some research objectives (here, you will
propose these objectives yourself). You should incorporate proper statistical analysis
methods to fulfil these objectives.

At your second task at the organisation, they have asked you to select some indicators
(using the data sources above) which you believe tell a significant story, and to produce a
single-screen interactive dashboard to present this data. For example, it could be to
compare the trade situation of the least developed countries with developed countries.
Your dashboard is to be made publicly available on their website, so you should consider
how you can present the data to a general audience who may not have existing expertise in
the subject you choose.

Task 1: Statistical Analysis (60% of Total Mark)

The requirements for the proposed statistical analysis are:

1. Define research objectives based on the dataset. For instance, to compare the trade
situation of the least developed countries with developed countries.
2. Based on the objectives, select a sample of countries. Sample should be at least 10
suitable countries of your choice.
3. Choose a set of indicators according to the objectives with at least 10 years of data.
4. Start to complete the following tasks. Also, present and interpret your findings and
results in the report as much as you can and show the R analytics steps.

4.1. Do a comprehensive descriptive statistical analysis (e.g., Mean, Median, Mode,

Standard deviation, Skewness and Kurtosis) on the data.

4.2. Do a correlation analysis for the indicators and evaluate the results in the context
of your stated objectives.

4.3. As a researcher, define at least two hypotheses testing related to the objectives
and test them.

4.4. Do regression analysis. Explain why the selected regression techniques are
appropriate for the selected variables and defined objectives and show if you’ve
found any similar research in the literature.

3
 4.5. Do time series analysis. Explain why the selected techniques are better for the
defined objectives and show if you’ve found any similar research in the literature.

Task 2: Interactive Dashboard Design (40% of Total Mark)

The requirements for the proposed dashboard are:

1. Clearly define the objectives of the dashboard based on the dataset you have selected.
2. Based on the objectives, select at least 10 suitable countries of your choice.
3. Produce a single-screen interactive dashboard of at least 10 countries’ data.
4. Clear, effective presentation of all factors in a coherent, intuitively comprehensive form,
reflecting the objectives you have set for your dashboard.
5. A design applicable to the full range of countries presented in the dataset without
modification to the dashboard form or structure. (i.e., the dashboard should support a
side- by-side comparison of multiple countries and/or financial years).

Alongside the dashboard design, you should provide a full report which summarises:

• The objectives you have defined for your dashboard, indicating clearly what your
planned solution will communicate to your audience.
• The data visualisation principles which have informed your dashboard design, with
reference to literature and best practice in data visualisation.
• The steps you have taken to pre-process and prepare the data.
• An overview of your design with a full justification of the design rationale

For extra credit you should also implement the following advanced features in your
dashboard design:

• Use of DAX
• Use of relationships in your data model
• Use of hierarchies, grouping or binning
• Use of in-built Power BI tools

To receive extra credit, these features must be fully documented in your accompanying
report.

Remarks:
1. Use similar datasets and objectives for both tasks. Although. if you prefer you can select
different objectives for each task.
2. You must use R programming language for the entire statistical analysis part (task 1).
3. Based on the data that you have downloaded from the given sources, you can mix
different data sets / variables to make your own data set in a meaningful and correct
format.
4. You must Include the screenshot of the final dashboard in the pdf report (task 2).

4
Assessed intended learning outcomes:

On successful completion of this assessment, you will be able to:

A- Knowledge and Understanding

1. Analyse a data science project to devise a structure for its implementation, analysis,
and evaluation, justifying any decisions made.
2. Critically assess the relative strengths and uses of a range of statistical analysis
techniques (including t-tests, ANOVA, various regression models and categorical data
analysis, test of hypothesis, and time series analysis).
3. Present and visualise the statistical results, analysing key findings.
4. Evaluate the quality of graphs according to their expressiveness and effectiveness.

B- Practical, Professional or Subject Specific Skills

1. Understand the history and context of data science ethics, skills, challenges, and
methodologies the term implies.
2. Will learn how to work with a real-world dataset that possibly is not in your domain
expertise, and you don’t have prior knowledge and understanding of that field.
3. Develop skills in presenting quantitative data using appropriate displays, tabulations,
and summaries.
4. Understand the nature of sampling variation and the role of statistical methods in
developing and testing hypotheses.
5. Select and use appropriate statistical methods in the analysis of complex datasets.
6. Present findings based on statistical analysis in a clear, concise, and understandable
manner.
7. Select the proper visualization methods for a given data analysis and presentation
problem.

C- Transferable Skills and other Attributes

5. Technical report writing.

6. Ability to use tools and techniques for statistical analysis.
7. Presenting data in a manner accessible to non-technical stakeholders.
8. Data Science Ethics, Information governance, information Literacy and Data Protection

Module Aims
The module is focused on the underpinning knowledge and practical skills needed for
working within the data sciences industry.

Word count/ duration (if applicable)

Your assessment should be between 6000 to 8000 words (between 30 to 40 pages).

5
 Feedback arrangements

You can expect to receive individual feedback in the form of an annotated marking
matrix with specific comments for each section, general comments for the work and up
to 3 specific areas for improvement.

Support arrangements
You can obtain support for this assessment by contacting Dr Kaveh Kiani or Nathan
Topping for the technical aspects of the module. Further support can be obtained from
the university as follows:
askUS
The University offers a range of support services for students through askUS.
Good Academic Conduct and Academic Misconduct
Students are expected to learn and demonstrate skills associated with good academic
conduct (academic integrity). Good academic conduct includes the use of clear and correct
referencing of source materials. Here is a link to where you can find out more about the
skills which students require http://www.salford.ac.uk/skills-for-learning.
Academic Misconduct is an action which may give you an unfair advantage in your
academic work. This includes plagiarism, asking someone else to write your assessment
for you or taking notes into an exam. The University takes all forms of academic
misconduct seriously. You can find out how to avoid academic misconduct here
https://www.salford.ac.uk/skills-for-learning.
Assessment Information
If you have any questions about assessment rules, you can find out more here.
Personal Mitigating Circumstances
If personal mitigating circumstances may have affected your ability to complete this
assessment, you can find more information about the personal mitigating circumstances
procedure here.
Personal Tutor/Student Progression Administrator
If you have any concerns about your studies, contact your Personal Tutor or your
Student Progression Administrator.

Assessment Criteria
It would be best to look at the assessment criteria to determine what we are explicitly
looking at during the assessment.

In Year Retrieval Scheme

Your assessment is not eligible for in year retrieval. If you are eligible for this scheme,
you will be contacted shortly after the feedback deadline.

Reassessment
If you fail your assessment and are eligible for reassessment, you will be allowed to re-do
the assignment based on the feedback given. The submission for this will be based on
university’s reassessment calendar and routines.

6
Assessment Rubric

Scale Mark Rank Statistical Analysis Description Data Visualization Description

• Aim and objectives have been defined and • Comprehensive practice in visualisation
adequately explained. techniques.
• More than minimum sample size has been • Consideration of comparative analysis
used, and the approach for selecting this presented within the visual representation all
sample has been justified. data ranges and scales.
• Advanced consideration of data preparation. • Expanded definition of a common perceptual
• Besides general preparation, handling missing model and justification of this in a cognitive
data and outlier detection algorithms have context.
been utilized. • Detailed justification of approach taken,
• In depth descriptive statistical analysis has principles adopted, and assumptions made
been provided. about the representational paradigm.
• Demonstrating comprehensively the R • Comprehensive representation of data using
analysis steps. tailored visualisations that extend and refine
Outstanding

➢ Outstanding correlation, regression and time the basic tool functionality.

90 - 100

series analysis has been done (2 advance Reg • Detailed and thorough definition and
and 2 TS models) justification of bespoke data representations
➢ Correct predictions have been made based that define appropriate data-centric displays
on 4 models. and features.
➢ Outstanding comparative analysis of the • Clear and consistent format and layout with a
hypothesis testing and more than 2 test of reasoned and justified perceptive and
hypothesis has been included. cognitive feature set throughout.
➢ All the results consist of highly precise and • A highly objective focused representation
well-explained statements for both technical that presents all evidence and draws the
and non-technical audiences. conclusion for the task objective.
• A detailed and thorough critical review of the
proposed data visualization with
consideration of the task and matching of the
Distinction

form presented to the task objectives.

• Use of the additional features mentioned in
the brief.
• Aim and objectives have been defined and • Strong practice in visualisation techniques.
adequately explained. • Consideration of comparative analysis
• More than minimum sample size has been presented within the visual representation all
used, and the approach for selecting this data ranges and scales.
sample has been justified. • Expanded definition of a common perceptual
• Advanced consideration of data preparation. model and justification of this in a cognitive
• Besides general preparation, handling missing context.
data and outlier detection algorithms have • Justification of approach taken, principles
been utilized. adopted, and assumptions made about the
• In depth descriptive statistical analysis has representational paradigm.
been provided. • Comprehensive representation of data using
• Demonstrating comprehensively the R tailored visualisations that extend and refine
analysis steps. the basic tool functionality.
Excellent

80 - 89

o Correct correlation, regression and time • Definition and justification of bespoke data
series analysis has been done (2 advance Reg representations that define appropriate data-
and 1 TS models) centric displays and features.
o Correct predictions have been made based • Clear format and layout with a justified
on 3 models. perceptive and cognitive feature set
o Excellent comparative analysis of the throughout.
hypothesis testing and more than 2 test of • An objective focused representation that
hypothesis has been included. presents all evidence and draws the
o All the results consist of precise and well- conclusion for the task objective.
explained statements for both technical and • A critical review of the proposed data
non-technical audiences. visualization with consideration of the task
and matching of the form presented to the
task objectives.
• Use of the additional features mentioned in
the brief.

Assessment Information/Brief
 Scale Mark Rank Statistical Analysis Description Data Visualization Description
• Aim and objectives have been defined and • Detailed consideration of comparative
adequately explained. analysis presented within the visual
• More than minimum sample size has been representation for a small selection of
used, and the approach for selecting this countries.
sample has been justified. • Detailed consideration of a common
• Advanced consideration of data preparation. perceptual model and justification of this in a
• Besides general preparation, handling missing cognitive context.
data and outlier detection algorithms have • Refined representation of data using tailored
been utilized. representational forms that extend and
• In depth descriptive statistical analysis has refine the basic offering of the packages
Very Good

been provided. considered.

70 - 79

• Demonstrating comprehensively the R • Thorough task-driven representation focuses

analysis steps. on the overall task objectives and has
❖ Correct correlation, regression and time selected and presented the data to meet
series analysis has been done (1 advance Reg these objectives.
and 1 TS models) • A developed justification of approach based
❖ Correct predictions have been made based on human perception and cognition for
on 2 models. functional elements but without
❖ Very good comparative analysis of the consideration of scheme as a whole.
hypothesis testing and more than 2 test of • Specific referenced evidence to support
hypothesis has been included. design decisions and overall coherence of the
❖ All the results consist of good statements for visual layout.
both technical and non-technical audiences.
• Aim and objectives have been defined and • Some consideration of comparative analysis
explained. presented within the visual representation
• More than minimum sample size has been for a small selection of countries.
used. • Some consideration of a common perceptual
• Consideration of data preparation. model and justification of this in a cognitive
• Besides general preparation, handling missing context.
data and outlier detection algorithms have • Representation of data using tailored
been utilized. representational forms that extend and
• Good descriptive statistical analysis has been refine the basic offering of the packages
provided. considered.
Merit
60 - 69
Good

• Demonstrating comprehensively the R • Task-driven representation focuses on the

analysis steps. overall task objectives and has selected and
• Correct correlation, regression and time presented the data to meet these objectives.
series analysis has been done (2 Reg and 1 TS • Justification of approach based on human
models) perception and cognition for functional
• Correct predictions have been made based elements but without consideration of
on 2 models. scheme as a whole.
• Good comparative analysis of the hypothesis • Some referenced evidence to support design
testing and 2 test of hypothesis has been decisions and overall coherence of the visual
included. layout
• All the results consist of good statements.
• Aim and objectives have been defined. • Functional representation of raw data based
• Minimum sample size has been used. on standard representations with some
• Minimum data preparation. modification of the attributes.
• Satisfactory descriptive statistical analysis has • Indirect representation of data with minimal
been provided. analysis or pre-preparation.
• Demonstrating R analysis steps. • Some justification of approaches and
• Correct correlation, regression and time principles applied.
series analysis has been done (1 Reg and 1 TS • A consistent but basic report that shows how
Satisfactory

models) general principles and approaches have been

50 - 59

Pass

• 2 tests of hypothesis has been included. used to define a coherent presentation.

• Some results have explanations. • Task focused presentation that considers the
objectives set but does not fully justify.
• A consistent report that considers the
functional layout and data representation
without justification against human
perception and/or cognition.
• Some supporting research of visual form
decisions based on unstructured and
validated web-based presentations of data
forms

Assessment Information/Brief
 Scale Mark Rank Statistical Analysis Description Data Visualization Description
• Aim and objectives have not been defined • Functional representation of raw data based
properly. on standard representations with minimal
• Less than minimum sample size has been modification of the attributes
used. • Indirect representation of data with minimal
• Minimum data preparation. analysis or pre-preparation.
• Unsatisfactory descriptive statistical analysis • Little justification of approaches and
has been provided. principles applied.
• Unsatisfactory R analysis steps. • A consistent but basic report that shows how
Unsatisfactory

• Unsatisfactory correlation, regression and general principles and approaches have been
40- 49

time series analysis has been done (some of used to define a coherent presentation.
the models are wrong) • Task focused presentation that considers the
• Requested test of hypothesis has not been objectives set but does not justify rationally.
included. • A basic report that considers the functional
• Results have not been explained. layout and data representation without
justification against human perception
and/or cognition.
• Little or no supporting research of visual form
decisions based on unstructured and
validated web-based presentations of data
forms
• Aim and objectives have not been defined. • Less than the minimum number of countries
Fail

• Less than minimum sample size has been have been visualised
used. • Functional data representations using basic
• Minimum data preparation. forms with minimum modification
• Inadequate statistical analysis has not been • Inconstant/uncoherent report that has little
Inadequate

provided. or no constancy of form or intent.

30 - 39

• Inadequate R analysis steps. • Little attempt to address the task focus or set
• Inadequate correlation, regression and time clear objectives for the dashboard
series analysis has been done (Most of the
models are wrong).
• Requested test of hypothesis has not been
included.
• Results have not been explained.
The submitted assessment report diverges significantly from the provided assessment brief,
displaying a significant lack of alignment between the given guidelines and the content presented.
Poor / Very poor /
Extremely poor

The report appears to lack the even basic elements outlined in the brief, deviating from the
specified criteria and objectives.
0 - 29

Assessment Information/Brief