When we refer to surveys, we are typically talking about a process that provides a snapshot from which you’re able to draw conclusions about a single time, place and group of people.
While this is usually sufficient, there may be occasions when you’re more interested in finding out how the people you surveyed are changing. In such a scenario, a single survey or snapshot is no longer enough, and you will need to repeat your surveys in order to track and see how things change over time.
This is where the use of longitudinal surveys or the longitudinal study as it’s otherwise known, where you follow the same group of respondents over an extended period of time, for weeks, months, or even years, can be extremely beneficial.
Types of longitudinal studies
When it comes to longitudinal survey research there are three main types of studies.
A cohort study is one that involves selecting a sample group to study who share the same common characteristics based on a specific event such as birth, geographic location or historical experience. Surveys of this cohort group are then conducted at various intervals usually over many years to study the data and see what factors may affect the outcomes and achievements of different individuals within that group.
A good example, is the well renowned UK birth cohorts, which follow groups of people born during the same time period and revisits them through surveys at various stages throughout their life.
In contrast to cohort studies, panel surveys involve sampling a cross-section of individuals to study across a much broader range of issues. Repeated surveys will be conducted at various intervals over time, which can be for as long as, or a much shorter period than cohort studies, in order to gather data from subjects and see how their opinions may have changed.
A retrospective study or historic cohort study as it’s otherwise known, is one that examines what has happened in the past in order to get answers.
This type of research looks to compare individuals that are similar in many ways but differ in terms of some key characteristics – such as female nurses who smoke and those who do not – in order to study a particular outcome, such as those that go on to develop lung cancer.
Data on relevant events for each individual (the form and time of exposure to a factor, the latent period and the time of any subsequent occurrence of the outcome) are then collected from existing records and immediately analysed to determine the relative risk of the cohort (those who smoke) compared to the control group (those that don’t smoke).
What is a longitudinal survey?
The longitudinal survey can be defined as a research process that involves repeated observations of the same groups of people, whether that’s employees, a specific group of customers, or another audience group to see if their opinions have changed since you first collected their data. Through repeat surveys, which in this example could include staff surveys, customer surveys and healthcare surveys if you were targeting patients, you could regularly check the views of these groups over a short or much longer period of time.
Looking at the three main longitudinal study types we outlined earlier, repeated surveys could be used to support them in the following ways.
For cohort research such as birth cohort studies, repeated surveys alongside interviews could be used to collect information ranging from education and employment, family and parenting to physical and mental health, and social attitudes, to examine which factors can most greatly affect outcomes and achievements in later life.
When it comes to panel studies, the use of repeated surveys could be beneficial for examining a much wider range of issues.
For example, if you worked in marketing and wanted to see how your customers might react to a new email newsletter design, you could send them the same concept evaluation survey twice. Your initial survey could be used to find out what they liked and didn’t like about your current design – you could even use this feedback to inform your new design. You could then issue them with a follow up survey after they’ve received your newly designed newsletter to see how their attitudes and opinions between the older and newer designs compare.
To support a retrospective longitudinal study, you could use a qualitative style survey with open-ended questions to gain a more in-depth picture of relevant events that happened in the past.
For example, if you were doing research on a group to investigate the factors that may have potentially contributed to some of them developing mental health issues, besides checking back through their medical records, you could run patient feedback surveys to get a bit more background information about them. With open-ended survey questions, you could retrospectively delve into areas including prior lifestyle and social factors that may have contributed towards the condition they are suffering from today.
Advantages and disadvantages of longitudinal studies
When considering different types of surveys there’s plenty to think about before starting a longitudinal survey, including an overview of the strengths but also any potential weaknesses, as mitigation will be essential if you’re to get the best out of this research approach.
The benefits of longitudinal study
Effectively examine changes over time
In contrast to many other types of research, longitudinal studies and surveys offer unique insight that might not be possible using any other research method. And because the longitudinal method allows researchers to look at changes over time, it’s particularly useful for studying lifespan and development issues. With the longitudinal approach, researchers can examine how certain things change at different points in life and explore some of the reasons why these developmental shifts take place.
They can provide clear focus and validity
With the ability to track changes over a period of time, which allows us to observe how a certain set of circumstances or an end state has come to be, longitudinal research can enable much greater clarity. In addition, even if people can’t remember past events, this problem can be easily addressed if they are recorded, which will ensure a high level of validity.
They are highly flexible
Compared to other research approaches longitudinal studies have been deemed to allow much greater flexibility. This means that their focus can be shifted, even while researchers are still collecting data.
Greater accuracy when observing changes
Given its reputation for being the best method for conducting research on developmental trends, and its ability to make the observation of changes more accurate, longitudinal research is seen as the go to option in various fields. For example, in medicine, longitudinal studies are used to discover predictors or indicators of certain diseases, while in advertising, they can be used to determine the impact that a campaign has made on the behaviour of consumers who have seen particular advertisements.
Problems with longitudinal studies
While there are plenty of good reasons to conduct a longitudinal study, there can be some weaknesses. So, it’s good to be aware of the possible disadvantages of longitudinal studies before you get stuck into your research.
They can take up huge amounts of time
Time is definitely a major drawback of any longitudinal study that you need to consider, as it will typically take a substantial amount of time to collect all the data that you require. It can also take an equally long period of time for you to gather your results and begin to make sense of the data.
They can be rather expensive
Given the huge amounts of time that longitudinal studies require, repeatedly observing and surveying the same subjects, they are often a lot more expensive than other research methods.
Participants may drop out over time
The risk with any research conducted over a significant period of time is that participants can sometimes drop out of a study, shrinking the sample size and decreasing the amount of data collected. The underlying reasons can be numerous, from illness and moving away from the area, to simply losing the motivation to continue participating.
Small sample sizes can threaten the validity of your research
If a lot of people drop out of your study it can mean that your final group no longer reflects your original representative sample, which can harm the validity of your research. The same could apply if your sample size was not large enough from the outset of your study.
Longitudinal survey design
While careful design should be a key element of any research study, the same should apply to a longitudinal survey, following the completion of your survey objectives.
With a survey design process in place offering some firm guidance about what you need to include and adhere too, it will help keep your survey on track and maximise your chances of meeting your overall objectives.
Here’s some key survey design considerations you need to be thinking about.
When it comes to your survey design, ensuring continuity throughout is one of the most important things you need to be adhering to, with your survey questions very much at the heart of this.
Research has shown that changing the way a question is asked can result in substantially different answers, even when it is aimed at the same people. So, if you’re to ensure the greatest accuracy when analysing your data, you need to be asking the same questions every time you repeat your survey.
It’s a fine balance, but you need to think carefully about how frequently you will run your survey and find the right equilibrium. If you conduct surveys too frequently, you risk wasting resources and time, since not enough time may have passed for any change to occur. You could also risk annoying respondents who are fed up with taking your surveys and may drop out of your study altogether. In contrast, if you don’t survey respondents enough you may miss out on some crucial data.
When you’re repeating surveys as part of your research, you also need to consider your survey sample size, to allow for the fact that not everyone who responded to your first survey is likely to respond to the second or third one. So, if you’re planning to run a longitudinal survey where you will survey the same group of people three times and you want at least 1000 respondents for your third survey, you’ll need to survey more than 1000 respondents in your first survey.
Test your survey first
While it’s important to ask the same questions in every survey you repeat, it’s also essential that your questions are generating the sorts of responses you were hoping for. So, it’s important that your question types and content are as good as they can be from the offset.
Testing your survey on colleagues or a beta group first can be an excellent way of checking this, enabling you to tweak your questions before you start your study and survey for real.