6047CEM - Designing and Planning your Research

Research Design (1)

What is Research Design?

The planning and structuring of your research investigation
Conceived to get answers to your research questions
Provides an outline of what you are planning to do
- i.e. from writing a hypothesis to the analysis of the data

Research Design (2)

What is a Research Question?

A statement of the problem you are researching
- presented in the form of a question
It will serve as the focus of your investigation
A good research question will:
- find the direction of the project
- narrow the scope of research
- define an investigation and set its boundaries

On the previous slide, I mentioned that the purpose of the research project is to obtain an answer to your research question or questions. But what do we mean by the term research question? Well, a research question is a statement of the problem that you are researching or investigating. This will often be presented in the form of a question, hence the term research question.

It will be defined to serve as a focus for your investigation. You may have a single research question, or you may have multiple. It will help you find the direction of your project, i.e. what is it that you are trying to get out of it. In some cases, it can also help narrow the scope of your research. For example, if you want to do a project on penetration testing, then an appropriate research question may be: can penetration testing aid in finding vulnerabilities on Android mobile devices? This has helped narrow the scope from just penetration testing in general, which can be applicable to multiple types of devices, i.e. mobile phones, tablets or internet of thing devices, to something more that is manageable, i.e. Android mobile phones.

Research Design (3)

Good Research Questions

Four essential characteristics for a research question:
- feasible
- clear
- significant
- and, ethical

This leads us nicely onto the concept of how we can go about forming a good research question. A good research question will follow four characteristics:

the question should be feasible, which means that it can be investigated in the confines of the project timeline and without an undue amount of effort or money
the question should be clear; this means that the question should be agreeable to the reader on what they keywords mean; it should not be ambiguous or confusing.
the question should be significant, whereby it is worth investigating because it will contribute towards some form of new knowledge, or building upon existing work already published.
finally, the question should be ethical, something that we have touched upon in a previous lecture.

Research Design (4)

Aims and Objectives

Your research question will influence the aims and objectives of your project
An aim will describe the main goal of your research project
- this can be the research question
Objectives are points that will help you answer the question(s)
- divides your aim/question into several smaller parts

Aims

Your aim should be made up of three parts, answering the following questions:
1. why is this research required?
2. what is the research about?
3. how are you going to go about doing it?
Aims are not limited to just one, similar to how you can have multiple research questions

Objectives

Objectives should be SMART:
- specific
- measurable
- achievable
- relevant
- and, time-bound

The design of your research question can help influence the overall aims and objectives of your project. The aims and objectives you design and state will help you meet answering your research question. The creation of an aim will set the main focal point of your research and will provide the readers with clarity as to what your project is all about. In some cases, the aim of your project may be the actual research question, and the objectives will help your project answer it. The purpose of an objective is to help you split your research question into several smaller parts, that can be made achievable.

Show the Next Fragment

When it comes to writing the aims of your project, you will want to consider the following questions: why is this research required? what is the research about? and, how are you going to go about doing it? To achieve your answers to these questions, you will want to write about each question as its own sentence or paragraph. You will want to ensure that you provide an answer to each one in the introduction of your report. The first question will provide the context of your research, whilst the second will describe the aim of your work and the third question providing an introduction to the objectives of your project.

The aims you provide may not be limited to just one. There may be one overarching aim that needs to be met, but this may split of to two or three aims. However, you must ensure that all aims are met. Whilst having more than one research aim is not disadvantageous, you may want to consider whether a single overarching aim will be better.

Show the Next Fragment

When it comes to creating the objectives of your research project, you will want to ensure they are SMART, and not the intellectual form. SMART is an acronym for: specific, measurable, achievable, relevant and time-bound. However, what do we mean by each of these? Well, let’s have a look at each one.

Specific is concerned with ensuring that there is no ambiguity in the action you are going to undertake. Therefore, you will want to ensure that the objectives are focused and well-defined. Measurable is concerned with how you will measure the progress of your project, and determine whether you have achieved the specific outcome required.

When we refer to an objective being achievable, we are concerned with whether you have the necessary support and resources to meet the objective. In this case, you may want to consider whether you have the necessary hardware or permissions to carry out your research.

Relevant is where you ensure that the objective is essential to the achievement of your research aim or question. You will want to ensure that you do not provide an objective that has no fundamental requirement for your project.

Finally, the last one is time-bound, whereby you need to be realistic on whether the objective can be met in the available time. You will only have the time of semester two to work upon your project, and you will also be studying and completing assignments for other modules during this time. Therefore, ensure any objectives you create can be met and are achievable within the realms of your other commitments.

Research Design (5)

Steps to Designing your Research

There are six steps to follow when designing your research:
1. Consider your aims and approach
2. Choose a type of research design
3. Identify your population and sampling method
4. Choosing your data collection methods
5. Plan your data collection procedures
6. Decide on your data analysis strategies

When it comes to designing your research, there are six steps you would potentially want to follow:

consider the aims and approach you will undertake for your project
choose the type of research design that would be most applicable
identify your population and sampling method
choose the methods in which you want to collect your data
plan the way in which your going to collect your data
decide on the necessary analysis strategies for your newly acquired data

Do not fret if you do not what each step would involve, as over the next few slides we shall look at each of these steps in detail.

Research Design (6)

Consider your Aims and Approach

Research should be driven by your aims and priorities
Start off by thinking about what you want to achieve
Decided whether your research is qualitative or quantitative
- qualitative allows the research to be more flexible and inductive
- quantitative tend to be more fixed and deductive
Possible to use a mixed approach combining both methods
- you can gain a more complete picture of the problem
- strengthen the credibility of your conclusions

Let’s begin with the first step, whereby you could consider your aims and the approach of your research. Before you can begin, you should already have a clear idea of what your project is and what you are undertaking. This should include a potential hypothesis or research questions in which you seek to answer.

There may be many ways in which you could go about answering your questions, and the research design should be driven by your aims and priorities. Therefore, you will want to start by thinking about what you wish to achieve from your project.

The first choice you should decide about your research is whether you want to follow a qualitative or quantitative approach. A qualitative approach allows for your research to be more flexible and inductive, which would allow for you to adjust your approach based on the findings from your research. A quantitative approach tends to be more fixed and deductive, with a collection of variables and a hypothesis clearly defined in advanced of the data being collected.

It would be possible to use a hybrid approach, whereby you would consider using a mixture of qualitative and quantitative research for your project. This will enable you to gain a more complete picture of the problem you are tackling and strengthen the credibility of the conclusions you draw.

Research Design (7)

Choosing a Type of Research Design

Both quantitative and qualitative approaches have several research designs
Each type of research will provide a framework to shape your research

Quantitative Research Design

Experimental
- test casual relationships
- subjects randomly assigned to groups
- conducted in a controlled environment
Quasi-Experimental
- test casual relationships
- no random assignment to groups
- comparing outcomes of pre-existing groups
Correlational
- test whether variables are related and how strong
- variables are measured without influencing them
Descriptive
- used to describe characteristics, averages, trends etc.
- variables are measured without influencing them

Qualitative Research Design

Case Study
- detailed study of a specific subject
- data is collected using a variety of sources and methods
- focuses on gaining an understanding of the case
Grounded Theory
- aims to develop a theory inductively
- analyses qualitative data
Phenomenology
- aims to understand a phenomenon
- describes a participants lived experiences

Within both qualitative and quantitative research approaches, there are a variety of different types of research design. Each type of design provides a framework which can provide the overall shape of your research.

Show the Next Fragment

We shall first begin by looking at the types of quantitative research design. It can be split into four main types:

experimental
quasi-experimental
descriptive
and correlational

Experimental and quasi-experimental will enable you to test cause-and-effect relationships. Whilst descriptive and correlational research allows you to measure variables and describe the relationships that exist between them. You can see on the slide some purpose and characteristics for each type of quantitative research design.

With descriptive and correlational designs, you are able to get a clear picture of characteristics, trends and relationships as they exist in the real world. However, you will not be able to draw conclusions about the cause and effect. In this instance, experiments are the strongest way to test cause-and-effect relationships without the risk of other variables influencing the results. However, due to their controlled nature, they often do not always reflect how things may work in the real-world. Therefore, they are often most challenging and in some cases expensive to implement.

Show the Next Fragment

Qualitative research design on the other hand are less strictly defined. With this approach to research, you will be concerned about gaining a rich and detailed understanding of a specific context. With this type of research, you can often be more creative and flexible in the design of your research.

On the screen, you can see some common types that are associated with qualitative research. Whilst they may often have a similar approach to the method of collecting data, they focus on different aspects when it comes to analysing your data. For example, with a case-study approach, there would be a detailed study for a specific subject. This could be a place event or an organisation. The data often collected will be using a variety of different sources and methods and will focus upon achieving a holistic understanding of the case.

Research Design (8)

Identify your Population and Sampling Method

Clearly define who or what your research is focusing upon
State how you will go about choosing participants or subjects
A population is the entire group you want to draw conclusions about
A sample is a smaller group of individuals who you will collect data from

Defining a Population

Generally made up of anything you wish to study
- i.e. organisations, research paper, animals etc

Sampling Methodologies

With a narrow population, it is challenging to collect data from every participant
Sampling is a useful approach, performed in one of two approaches:
1. Probability:
  - sampling via random methods
  - used mainly in quantitative methods
  - enables a strong statistical inference about the population
2. Non-Probability:
  - sampling via a non-random method
  - used in both qualitative and quantitative methods
  - easier to achieve, but there is a greater risk of bias

The third stage is the identification of your population and the sampling method. Your research plan should clearly define who or what your research will be focusing upon, and how you may go about choosing participants or tools. For example, if you were looking at penetration testing internet of things devices, then you will want to consider which devices you are going to be using? You should not be limiting yourself to a single device; you will want to consider a broader range of devices, manufacturers and types of IoT devices. For example, you will not just want to perform a penetration test on a single light-bulb, you may want to consider plugs, temperature sensors, etc. and those from different manufacturers.

Show the Next Fragment

In research, a population is the entire group that you wish to draw a conclusion about, whilst a sample is considered to be a smaller group from which you will collect your data from. Going back to my example of IoT devices, you could consider your sample being a smart plug and light-bulb from the parent company. The parent company may have other IoT devices, but if your pen-testing highlights security issues in your two sample devices, then it could be that the suite of IoT devices by the parent company may have these issues.

A population can be made up of anything you want, it could be people, hardware, text, etc. The more precisely you are able to define your population, the easier it will be to gather a representative sample. It may be the case that your population is too large, and as such, you will want to take a smaller sample.

Show the Next Fragment

Therefore, to select a sample, there are generally two approaches: probability and non-probability sampling. The sampling method in which you will use will affect how confidently you can generalise your results to the population as a whole. Probability sampling is the most statistically valid option when it comes to choosing a sample from a population. However, it is often challenging to achieve unless you are dealing with a tiny and accessible population.

Therefore, for a practical reason, many studies use a non-probability sampling method; however, it is important to be aware of the limitations with this type of sampling and consider any potential biases. You should always make an effort to gather a sample that is representative of the population.

I should also point out that in some cases of qualitative research designs, sampling may not be relevant at all.

Research Design (9)

Choosing your Data Collection Methods

A method of directly measuring variables and gathering information
Provides you first-hand knowledge and original insights into your research problem
You may do one data collection method, or multiple:
- e.g. surveys, observations or media

Primary Data

This is data collected by the researcher
Often referred to as “real-time” data
- the data is collected when the research is being undertaken

Secondary Data

This is data collected from other researchers
- i.e. data sets from Kaggle, or other research papers
You provide your own analysis on the raw data
- this can be done to answer new research questions not originally addressed from the study
Expand the scope of your research
- access larger and more varied data samples
There is a lack of control over the variables measured or how they were measured

The fourth stage of research design is choosing a methodology to collect your data. The data collection method is a way of being able to directly measure variables and gather information. This will enable you to gain first-hand knowledge and original insights into the research problem. You may want to choose a single data collection method, or you may want to use several methods in your project.

Show the Next Fragment

Primary data is considered to be data that has been collected by yourself. Sometimes, this can be referred to as “real-time” data, as the data has been collected at the time the research project has been undertaken. The data collected will be under the control of you, the researcher.

Show the Next Fragment

Somebody else often collects secondary data, generally in the past. Using secondary data is easy as you are not required to collect the data yourself. However, secondary data is not always relevant to the project you will be undertaking as it was often collected with a different research question in mind. There may also be concerns about how the data has been collected.

Research Design (10)

Plan your Data Collection Procedures

Think about how you will execute the methods of data collection
You will want to collect data that is consistent, unbiased and accurate
Have a systematic approach to your qualitative research
- define your variables and make sure your measurements are valid and reliable
Ensure that the results collected are reliable and valid
Create a data management plan
- how will you store the data?
- how will it be organised for easy retrieval?

The fifth stage is the planning of how you will collect your data. You will want to plan how you will use a decided data collection method to collect data that is consistent, accurate and not biased. This is especially important when it comes to quantitative research, as you will need to precisely define your control variables and ensure that the measurements obtained are reliable and valid.

You will want to ensure that any results obtained are reliable and valid. In this instance, we refer to reliability as meaning that the results can be consistently reproduced, while validity means that you are actually measuring the result you are interested in. For valid and reliable results, your measurement materials should be thoroughly researched and carefully designed. You will want to plan your procedures to make sure you can carry out the same steps in the same way for each experiment you undertake.

For example, if you are going to be pen-testing multiple IoT devices, then you want to ensure you take the same rigorous for each device you are testing. This will mean ensuring they are set up all in the same method (or as close as possible), performing the same pen-testing attacks, with the same variables, etc.

You will also want to consider creating a data management plan, especially if you are collecting sensitive data. How will this data be stored? Where will it be stored? How will you ensure external eyes will not make contact with sensitive data? You will also want to consider how you organise your data, as keeping it well-organised will assist you when it comes to analysing it.

Research Design (11)

Decide on your Data Analysis Strategies

Raw data will not answer your research question
You will need to consider analysing the data collected

Qualitative Data Analysis

Data will often be dense with information and ideas
Instead of summing it up via numbers, you’ll have to go through the data in detail
- i.e. interpret its meaning and identification of any patterns
You may also need to extract relevant parts of your research question
Common approaches towards this type of analysis:
- thematic: focuses on the content of the data and involves coding/organising the data to identify key themes
- discourse: puts the data into context by analysing different levels of communication
  - i.e. language, tone and structure etc.

Quantitative Data Analysis

Some form of statistical analysis will be performed
- i.e. summary of the sample data, estimations and test hypotheses
Descriptive statistics can be used to summarise sample data by looking for the:
- distribution of your data
- central tendency of the data
- variability of the data
Inferential statistics can be used to:
- estimate the population based on the sample data
- test hypotheses for a relationship between variables
The choice of test employed depends on the aspects of your research design

The final stage of your research design is making a decision on how you will analyse the data. The data you collect is considered to be raw, and it will not enable you to answer your research question as it stands. Therefore, some form of analysis will need to be undertaken.

Show the Next Fragment

If the research you have undertaken is qualitative, then the data you collect will often be dense with information and possibly ideas. Instead of being able to present the information in the form of numbers, you will need to go through the data in detail and interpret the meanings and identify any patterns from within it. You will extract parts of the data that are most relevant to your research question.

There are two common approaches to undertaking this: thematic and discourse analysis. A thematic analysis focuses upon the content of the data and will involve coding and organising the data to identify key themes. Whereas discourse analysis is focused upon putting the data into context. It will involve analysing different levels of communication, this may be the language, structure or tone, etc.

** Click Next Slide**

Analysing data collected from a quantitative research project is a little different. In this form of research, you will likely use some form of statistical analysis. Using statistics means that you can summarise your data into numbers, you can make estimates and test your hypothesis. There are generally two forms of statistical analysis that can be performed with this type of research: descriptive and inferential statistics.

Descriptive statistics are concerned with summarising the data in terms of its distribution, the mean or central tendency and the variability of the data, i.e. the standard deviation. On the other hand, inferential statistics are concerned with making estimates about the population based upon your sample data; it can also test your hypothesis about a relationship between given variables.

The type of tests that you choose and depend are based upon the various aspects of your research design. This will include the type of variables that you are dealing with and the distribution of the data.

Designing and Planning your Research

Dr Ian Cornelius

Hello

Research Design

Research Design (1)

What is Research Design?

Research Design (2)

What is a Research Question?

Research Design (3)

Good Research Questions

Research Design (4)

Aims and Objectives

Aims

Objectives

Research Design (5)

Steps to Designing your Research

Research Design (6)

Consider your Aims and Approach

Research Design (7)

Choosing a Type of Research Design

Quantitative Research Design

Qualitative Research Design

Research Design (8)

Identify your Population and Sampling Method

Defining a Population

Sampling Methodologies

Research Design (9)

Choosing your Data Collection Methods

Primary Data

Secondary Data

Research Design (10)

Plan your Data Collection Procedures

Research Design (11)

Decide on your Data Analysis Strategies

Qualitative Data Analysis

Quantitative Data Analysis

Goodbye

Goodbye (1)

Questions and Support