Skip to Main Content
It looks like you're using Internet Explorer 11 or older. This website works best with modern browsers such as the latest versions of Chrome, Firefox, Safari, and Edge. If you continue with this browser, you may see unexpected results.
Logo for the SkillsCentre

Dissertations and research projects

Online study guides for every stage of your research project, from planning to writing up. Also includes advice on writing a remote dissertation while social distancing measures are in place.


Through quantitative research we seek to understand the relationships between variables. A variable will be a characteristic, value, attribute or behaviour that is of interest to the researcher. Some variables can be simple to measure, for example, height and weight. By contrast, others such as self-esteem or socio-economic status are more complex and therefore harder to measure. This is why it is important to operationalise your variables.

This essentially means being very clear about the way in which variables will be defined and measured in your study; this lends credibility to your methodology and helps the replicability of your research. It is important that you are detailed in your operational definition of any given variable because another researcher may define that variable differently from you. To illustrate, if a study examined memory ability, the researcher would specify exactly how this measure was generated: was it the number of words recalled in 60 seconds after reading a passage of text? Was it details about a picture? Defining your variables is an important of the research process as this will affect the reliability and validity of your study.

is the variable changed or manipulated by the researcher. Research generally seeks to establish whether the independent variable has an affect or influences the dependent variable in some way; this may be through a causal or non-causal relationship.

is the variable that the researcher is trying to predict or explain through understanding its relationship with the independent variable. For example, if a researcher wants to establish if drinking coffee aids sporting performance, your independent variable would be the amount of coffee consumed (no coffee/1 cup/3 cups) and the dependent variable would be some operational definition of sporting performance (amount of weight lifted/vertical jump height/time taken to sprint 100m).

is a variable that affects the strength of the relationship between the independent and dependent variable. For example, if you looked at the relationship between personality similarity in friendships (independent variable) and perceived friendship satisfaction (dependent variable), it might be that age is a moderating variable – e.g. the older you are, the weaker the relationship between personality similarity in a friendship and associated satisfaction with that friendship. From this you could make the tentative suggestion that similarity in personality becomes less important in a satisfying relationship as we become older. 

is a variable that helps to explain the relationship between the independent and dependent variables. Consider the example above, we might discover that the number of shared activities also contributes to perceived friendship satisfaction. We could then remove this from our analysis and find that the relationship between personality similarity in friendships and perceived satisfaction in a friendship disappears - this would suggest that the relationship was mediated by the variable shared interests.

is any variable that is not the independent variable but may affect the results of the experiment. Examples can include; aspects of the environment (temperature/noise/lighting); differences between participants (mood/intellect/concentration); and experimenter effects (clues in an experiment which may convey that purpose of the research). It is important to minimise the influence of extraneous variables through the careful use of controls – for example, there are ways of minimising the effect of differences between participants through your experimental design (more on this later!)

Hypothesis testing

A hypothesis is a predictive statement that can be tested through the collection of data. The data can be analysed and can either provide support for, or help to reject, a hypothesis; this in turn should allow a researcher to draw some conclusions about what they are investigating.

Null and alternative hypotheses

Hypothesis are classified by the way they describe the expected association/difference between variables. When we test our hypothesis/hypotheses it is important to remember we are testing it against the assumption that there isn’t an association/difference between the independent and dependent variables: we call this the null hypothesis. By testing this assumption, statistical tests can estimate how likely it is that any observed association/difference between variables is due to chance.


In addition to the null hypothesis we also have the alternative hypothesis. This hypothesis states that there is an association/difference between groups; this cannot be tested directly but can be accepted by rejecting the null hypothesis. This is achieved through statistical tests that can help to demonstrate that any observed association/differences are not due to chance. Once this is established, we can accept our alternative hypothesis and start to draw conclusions from our data.


Hypotheses can either be one-tailed or two-tailed:

  • One-tailed hypothesis –specifies the direction of the predicted association between the independent and dependent variable. For example, the higher an individual’s educational level, the more books they will read in a one-year period.

  • Two-tailed hypothesis – does not specify the direction of the predicted association between variables; only that an association exists. For example, there will a be difference in the number of books read in a one-year period, dependent on the level of an individual’s education.

Key things to remember when writing your hypothesis/hypotheses:

  • Your hypothesis should always be written as a statement and before any data are collected.

  • It should be simple and specific; include the variables, using concise operational definitions, and the predicted relationship between these variables. If you have several predictor (independent) variables it would be better to write several simple hypotheses – think one predictor and one outcome variable.

  • Always keep your language clear and focused.
Find out more on hypotheses (including a discussion of  Type 1 and 2 errors and effect size).

Reliability and validity

It is important that you show rigour within your research. This means demonstrating that you have given careful consideration to how you can enhance the quality of your research project. Within quantitative research this is achieved through examining reliability and validity.


  • Reliability – is a measure of how consistent, dependable and repeatable something is.
  • Validity – is the extent to which research measures the concept that it was designed to measure.


For example, if you had some scales that were always weighed an object as 5kg lighter than it actually is, this would be an example of a measure that was very reliable but not valid: the scales will always give you a consistent measure of weight, but this measure is not accurate.


There are several different types of reliability and validity that you should consider when planning, conducting and writing up your research project. For more information on the different types of reliability and validity have a look at the recommendations below:

  • Designing and Doing Survey Research (Andres, 2012) – see Chapter 7.
  • Quantitative Health Research Issues and Methods (Curtis & Drennan, 2013) – see Chapter 16.
  • Research Methods in Psychology (Howitt & Cramer, 2017) – see Chapter 16.

Approaches to quantitative research

Non-experimental research designs do not seek to establish cause and effect relationships. This is because the researcher does not manipulate the independent variable(s) to measure any effects on the dependent variable(s). Instead, researchers may use this type of design to begin exploring a topic where there is little current understanding, or to investigate the relationship between two (or more) variables.

  • Descriptive: these research designs help to understand the current state of a phenomenon and are often used when not much is known about a topic. Variables are not controlled, and data tends to be collected through observation or surveys. An example of this might be an investigation into the preferred news sources of 13-18-year olds.

  • Correlationalthese designs measure a relationship between two variables that are not controlled. As such, correlational designs cannot establish cause and effect – always remember correlation does not imply causation! This approach can be useful when there is a suspected relationship between variables, but it would be impractical or unethical to manipulate one of those variables. For example, you might hypothesize

True experiments seek to establish cause and effect relationships between a group of variables. Researchers control for all variables except for the variable(s) being manipulated, to establish its effect on the dependent variable.

These are similar to true experiments: the aim is to establish cause and effect relationships. Crucially however, assignment to groups is not random. This type of design is often used when it is not possible for the researcher to randomly assign participants to groups because they are interested in understanding a particular phenomenon in relation to naturally occurring differences between groups – an example of this could be an experiment where a researcher is interested in examining whether the effect of coffee consumption on sleep differs depending on age. In this example, it is impossible for the researcher to manipulate the age of participants, so instead group assignment would be made based on predetermined criteria e.g. under 40, 40+. As this assignment cannot be random this would be a quasi-experiment.

Between-subjects designs involve the assignments of participants to one of two (or more) conditions, with each participant experiencing only that condition. In its simplest form, a between-subjects design requires a control condition and a treatment condition. If the results of an experiment differ greatly between conditions, then it can be assumed that this due to the effect of the intervention or manipulation that has been applied in the treatment condition. To help minimise the affect of extraneous variables that might impact differences between the groups (and increase the likelihood that observed differences are due to the effect of the independent variable), participants in the control and treatment conditions might be matched for relevant characteristics.

For example, in an experiment to assess the effectiveness of two training programmes in improving athletic performance, participants might be matched for some key measures of fitness such as 100m sprint time, maximum squat etc. This would enable researchers to be more confident that any changes to athletic performance in the participants between the two groups were likely due to the training programme they undertook, rather than natural, pre-exiting differences in athletic performance.

Sometimes referred to as repeated measures, this approach involves obtaining more than one measure from each participant in a study. This means that participants take part in both the control and treatment condition(s). The primary advantage of this is that participants act as their own control; you can be more confident that any observed differences result from the treatment condition rather than naturally occurring differences between the groups. One problem, however, is that of order effects (sometimes called practice effects). These effects may occur because conditions are applied one after the other and this can lead to changes in performance that are not the result of the treatment but instead reflect some effect of the previous condition that a participant has experienced. For example, improvements in performance could be due to learning/practise and a decline in performance could be due to fatigue over experiencing two (or more) experimental conditions back-to-back.

One way to account for this problem is to counterbalance the order of your conditions. To do this, a researcher would ensure that each condition in the experiment is experienced 1st for an equal number of participants:

10 participants experience condition A 1st and condition B 2nd
10 participants experience condition B 1st and condition A 2nd

Doing this helps to reduce the impact of order effects by ensuring that any effects are distributed evenly across all conditions.

Another option could be to create a long time between testing conditions to reduce any possible effects of learning and/or fatigue. It should be noted however, that creating distance between conditions isn’t always practical and it can be hard to know how long is sufficient to eliminate a potential order effect: this is particularly true for practice effects as it can be hard to accurately determine how long it takes for potential improvements in performance due to learning, to disappear.

The final type of design is used when a research design has one (or more) factors that is between subjects and one (or more) factors that is within subject. This is often used when for research that is looking at the effect on an intervention in relation to another factor that has a fixed effect.


For example, if a researcher was looking at the effectiveness of a drug for treating pain in those under 40 and over 40, age would be a between-subjects factor because a participant can’t be both under and over 40 at the same time. The drug that participants take would be the within-subjects factor. This type of design is particularly useful if you want to examine if the effect on an intervention is different dependent upon another factor. In the example above, it would be possible to establish if the effect of the drug was beneficial for all participants, or whether it was particularly effective/ineffective depending on the age of the participant – whether they were under or over 40.