Researchers use designs which are underpinned by certain philosophies. I would like to know from colleagues the appropriate paradigmatic stance that is consistent with the cross-sectional descriptive design.
Cheapness!
That is in comparison with other approaches ( eg a panel or cohort) but misses a lot due to lack of a time dimension. So it may be a useful if the thing you are studying is reactively unchanging and just want to know prevalence of the outcome variable.
Bur (seriously now) I do not buy the correspondence between design and philosophy
see some arguments here
https://www.researchgate.net/publication/256801634_The_practice_of_quantitative_methods
I do not believe that extensive design = quantification = positivism
and on comparing different types of design - Chapters 4 and 5 of
https://www.researchgate.net/publication/236671039_Epidemiology_An_Introduction
which is deliberately short and not too statistical
Book Epidemiology: An Introduction
Chapter The practice of quantitative methods
Dear Anthony,
As far as I understood, cross- sectional study design's research paradigm is positivism which depends on quantifiable observations/data for a specific point in time that lead themselves to make inferences.
It is relatively inexpensive and requires a little time. The cheapness depends on the area of interest and the feasibility to address an appropriate sample from a population.
Hope this would help.
Anthony,
Seems that you need no additional commentary on the relationship between research design and school of philosophy (e.g., positivism, mechanistic vs organic world views). However, the responses thus far have not included (perhaps deliberately) one aspect of research design, particularly in medical, behavioral and biological sciences. Designs that have lasted the test of time typically fit into two hierarchies: 1) level of evidence and 2) goals of science. Level of evidence is handled elsewhere in detail. However, few commentaries outside of undergraduate text books explain the progression of scientific endeavor. Though it is quite simple, it is often forgotten that the purpose of science is to a. describe b. predict and c. control. Within this paradigm, cross sectional studies almost without fail, satisfy the first level of scientific inquiry -- to describe. I am unsure whether this was helpful and my contribution is certainly not as lofty as some of the others, but, in scientific design, it is easy to miss the woods for the trees.
Dear Anthony:
It would be very useful for us if you define your word ''philosophy'', in this context, in order to get to know what you are interesting in.
Charles, for instance, left out one of the main purpose of science such as ''explain'', there is some of logical rout as : to describe, explain, predict and control or modify.
Some believe that the cross-sectional design is descriptive, the only thing you have to do in order to know that there a different view point is to read the following references:
David G. Kleinbaum, Lawrence L. Kupper & Hal Mogestern. Epidemiologic research: priniciples and quantitative methods. Van Nostrand Reinhold, 1982.
and,
Jennifer L Kelsey, W. Douglas Thompson & Alfred S. Evans. Methods in observational epidemiology. Oxford University Press, 1986. Chapter 8. Cross-sectional and other types of studies.
I hope this help.
A research design must be consistent with the research philosophy. cross-sectional design entails the collection of data on more than one case and at a single point in time in order to collect a body of quantitative or quantifiable data in connection with two or more variables. Cross-sectional designs are founded on positivism. the positivists believe in the fact that social reality is observable and the observed patterns through surveys can be generalized. Thus a cross-sectional descriptive study which is quantitative research in nature is suited for the positivist philosophy.
Two things rule out positivism for me as a workable philosophy in any science - physical , natural and social
1 you are not allowed to have directly unmeasurable entities - and yet science is full of them
2 explanation equals regularity - I do not believe that being told the reason why my train is late is that it is always late is an explanation. I want to know the mechanism and structures that are producing that lateness.
Additionally the human sciences deal with open systems: people are capable of monitoring themselves and changing. I do not believe that in fitting a model to your data collected by a cross-sectional design that this is an equation for all time and all places - the positivist's dream. It is perfectly possible to have an anti-positivist philosophy that uses quantitative designs when they are useful.
So for me practical adequacy is more important in meeting the tenets of some particular philosophy- but that of course is in some senses a philosophical claim!
For what has been called the Copernican moment in the philosophy of science see that explicates a non-positive and non-entirely social constructionist science
http://uberty.org/wp-content/uploads/2015/09/Roy_Bhaskar_A_Realist_Theory_of_Science.pdf
Dear all:
If we use the Chaos theory, our personal believes and action will be quite different in terms of philosophy posture. Thus, neither philosophy school of thinking has the revealed true, even more, the positivism theory establish that the ''cause'' comes first and thereafter we will have the effect, in a linear fashion and with only one factor involved at the time, but when tacking about the cross-sectional design most of the researchers say that one of its weakness is the difficult to establish the time relationship between the factors and the response variable. As we can see this is a clear conceptual contradiction, still unsolved for some people.
The real amazing situation for me is that some factor-response variable relationship are not linear, they could be quadratic or some other form, and some researchers force them to linear loosing a great portion of new knowledge.
As we all know in the Chaos theory the ''cause'' and response variable relationship is circular and multi factorial.
The learned lesson here, for me, is that there is not philosophy school of thinking capable to resolve our current lagoons of knowledge, and that the quantitative methods do not belong to a particular school of thinking.
Putting cost aside, a descriptive cross sectional study, is very good basis for a research to generate an hypothesis where there is none or one is not clear. because its brings out the prevalence (single point measurement , though others also define period prevalence) of a an item of interest, it become a good study design where information is decision making. The key for nay research design that a researcher chooses is to suit the purpose of the information required and its not a one size to fit them all.
- Badges
- Science topic
- Similar topics
- Medicine
- Epidemiology