Quantitative research is a crucial stage in virtually any research project, discovering the larger trends that can then be further examined.
Along with qualitative research, it is one of the twin pillars of understanding trends and user behaviour.
However, quantitative research is not a singular entity, there is not just one form of this analysis.
In this post, we example the four types of quantitive research, explaining them in simple, understandable terms and providing examples to explain their usage, benefits and potential limitations.
Quantitative data research relates to numbers and how you can group things together.
It is research that looks at a broad level and does not delve down into the individual.
Quantitative research would look to find which the most popular car make is, or how baby name choices have changed over time. It would not ascertain why car brand X is more popular among under 30s than car brand Y.
While different terms can be applied, the four different strands of quantitative research are:
We will taker these in order.
Descriptive quantitative research looks to describe the current status of a real-world phenomenon.
In this approach, the researcher does not start with a hypothesis, instead they gather data to then draw any conclusion or theory from.
Almost everyone will have been surveyed whereby descriptive information is being sought.
Examples might be acquiring data for shops a town centre would benefit from ‘which type of shop would you like to see in X town centre’ or attitude to switching to a four-term school system.
In each case, the researcher will collect the data and this can then be analysed and conclusions drawn and follow-ups designed.
It is essential that the initial data collection phase enables the correct choices to be recorded. -for instance, a survey asking about desirable new shops would be flawed if it only had a set list of optioned the failed to include choices that would be popular.
A correlational research project would explore a link between variables but without looking to apply cause and effect reasoning.
As with the descriptive market research, this is an observational form of quantitative research – indeed, sometimes it will be grouped with descriptive research as with both information is gathered without cause analysed.
Examples demonstrate that this is a form of research of which we are all familiar. Essentially, it can be the link between any two things.
The relationship between average hours of sleep and school attainment; the relationship between annual income and depression; the relationship between age and media consumption by type.
It would then be to draw out and further explore the legitimacy of these relationships – some could be purely coincidental. You could examine the relationship between length of surname and IQ, but would any apparent link have any legitimacy?
It is therefore important to select items to observe with care and then use the data as a starting point for further research rather than necessarily a fait accompli.
Quasi-experimental research is a powerful tool but one that needs skilful application to avoid potentially damaging and false correlations being suggested.
This form of research – also known as casual-comparative – looks to establish relationships between the variables, but it must also factor in other variables – both known and potentially unknown.
One example, and this is one that has caused past problems, is to look at education – for instance the impact of children taking multi vitamins on attainment. A link could be shown by the data – the link could be valid, but other factors could also be at play.
Do the children with diets have better overall diets? Do the parents giving their children multivitamins also typically take a more involved approach to their children’s education, perhaps even involving extra tutoring?
The researcher often has to make do with groups as they already exist – for instance the grouping of children by classes in school.
It is important for any research project to be clear that it is quasi-experimental rather than a pure experiment. This does not render the data invalid, but it factors in an acceptance that groups could not be chosen completely at random and so other variables will have an impact.
How much of an impact they have has to be ascertained or at least factored into any analysis.
Experimental research tests the relationship between variables but, unlike the quasi-experimental approach, this is done in a setting whereby there has been optimal variable control.
This approach can also be known as true experimentation. It can be difficult to set up but it leads to greater confidence and legitimacy in any end result.
The researcher will control – or at least attempt to control – every variable except for the one being manipulated (the independent variable).
Websites are often able to run this form of experimentation by serving up subtly different versions of pages to huge groups of users in a truly randomised way, with the results analysed.
For instance, font vs click-through rate could be analysed if all the content was the same and the only difference was that the two different font options were served to different users within a huge pool at random.
Being able to randomise to a large group is a key facet. In medicine, different treatments can be trialled this way – the effect of a new treatment plan on dementia for instance.
It hopefully will not surprise you to learn that there is no singular best approach for all circumstances – there is also the balance of quantitative and qualitative research to be factored in.
Instead, it is recommended to have a detailed consultation with experts in market research, outlining the areas you wish to explore and working with them to find a plan of best action.
A professional approach can ensure that valid insight is found, insight that can be acted upon and drive future policy and decisions for businesses, organisations or policy.
Quantitative market research is not automatically beneficial – it has to be the right questions asked in the right way, the right data collected and only valid conclusions drawn, or follow-up questions explored.
The proof of our quality is in our case studies and past clients. Please take some time to view our past work, this shows how we worked with clients to understand their needs, advise as appropriate and deliver the findings that could benefit their business.
We have won awards, received accreditation and have professional certification, for instance for data usage – you can find out more on this site.
We also have a bespoke verification programme called Acumonitor, this verifies all participants. We take every step to ensure you can be confident in the validity of the research and analysis we provide.
Acumonitor is an example of how we are actively looking to drive the standards of market research forward – you can read more and watch a short video that explains more.