A research design is the set of methods and procedures used in collecting and analysing measures of the variables specified in the research problem. The design of a study defines the study type (descriptive, correlation, semi- experimental, experimental, review, meta-analytic) and sub-type (e.g., descriptive-longitudinal case study, research problem, Hypothesis|hypothesesindependent and dependent variables, Design of experiments|experimental design, and, if applicable, data collection methods and a statistical analysis plan. Research design is the framework that has been created to find answers to research questions
Design types and sub-types
There are many ways to classify research designs, but sometimes the distinction is artificial and other times different designs are combined. Nonetheless, the list below offers a number of useful distinctions between possible research designs. A research design is an arrangement of conditions or collections.
Sometimes a distinction is made between "fixed" and "flexible" designs. In some cases, these types coincide with quantitative and qualitative research designs respectively, though this need not be the case. In fixed designs, the design of the study is fixed before the main stage of data collection takes place. Fixed designs are normally theory-driven; otherwise, it is impossible to know in advance which variables need to be controlled and measured. Often, these variables are measured quantitatively. Flexible designs allow for more freedom during the data collection process. One reason for using a flexible research design can be that the variable of interest is not quantitatively measurable, such as culture. In other cases, theory might not be available before one starts the research.
The choice of how to group participants depends on the research hypothesis and on how the participants are sampled. In a typical experimental study, there will be at least one "experimental" condition (e.g., "treatment") and one "control" condition ("no treatment"), but the appropriate method of grouping may depend on factors such as the duration of measurement phase and participant characteristics:
Confirmatory versus exploratory research
Confirmatory research tests a priori hypotheses — outcome predictions that are made before the measurement phase begins. Such a priori hypotheses are usually derived from a theory or the results of previous studies. The advantage of confirmatory research is that the result is more meaningful, in the sense that it is much harder to claim that a certain result is generalizable beyond the data set. The reason for this is that in confirmatory research, one ideally strives to reduce the probability of falsely reporting a coincidental result as meaningful. This probability is known as α-level or the probability of a type I error.
Exploratory research on the other hand seeks to generate a posteriori hypotheses by examining a data-set and looking for potential relations between variables. It is also possible to have an idea about a relation between variables but to lack knowledge of the direction and strength of the relation. If the researcher does not have any specific hypotheses beforehand, the study is exploratory with respect to the variables in question (although it might be confirmatory for others). The advantage of exploratory research is that it is easier to make new discoveries due to the less stringent methodological restrictions. Here, the researcher does not want to miss a potentially interesting relation and therefore aims to minimize the probability of rejecting a real effect or relation; this probability is sometimes referred to as β and the associated error is of type II. In other words, if the researcher simply wants to see whether some measured variables could be related, he would want to increase the chances of finding a significant result by lowering the threshold of what is deemed to be significant.
Sometimes, a researcher may conduct exploratory research but report it as if it had been confirmatory ('Hypothesizing After the Results are Known', HARKing—see Hypotheses suggested by the data); this is a questionable research practice bordering on fraud.
State problems versus process problems
A distinction can be made between state problems and process problems. State problems aim to answer what the state of a phenomenon is at a given time, while process problems deal with the change of phenomena over time. Examples of state problems are the level of mathematical skills of sixteen-year-old children or the level, computer skills of the elderly, the depression level of a person, etc. Examples of process problems are the development of mathematical skills from puberty to adulthood, the change in computer skills when people get older and how depression symptoms change during therapy.
State problems are easier to measure than process problems. State problems just require one measurement of the phenomena of interest, while process problems always require multiple measurements. Research designs such as repeated measurements and longitudinal study are needed to address process problems.
Examples of fixed designs
Experimental research designs
See also: Experiment
In an experimental design, the researcher actively tries to change the situation, circumstances, or experience of participants (manipulation), which may lead to a change in behavior or outcomes for the participants of the study. The researcher randomly assigns participants to different conditions, measures the variables of interest and tries to control for confounding variables. Therefore, experiments are often highly fixed even before the data collection starts.
In a good experimental design, a few things are of great importance. First of all, it is necessary to think of the best way to operationalize the variables that will be measured, as well as which statistical methods would be most appropriate to answer the research question. Thus, the researcher should consider what the expectations of the study are as well as how to analyse any potential results. Finally, in an experimental design the researcher must think of the practical limitations including the availability of participants as well as how representative the participants are to the target population. It is important to consider each of these factors before beginning the experiment. Additionally, many researchers employ power analysis before they conduct an experiment, in order to determine how large the sample must be to find an effect of a given size with a given design at the desired probability of making a Type I or Type II error.
Non-experimental research designs
Non-experimental research designs do not involve a manipulation of the situation, circumstances or experience of the participants. Non-experimental research designs can be broadly classified into three categories. First, in relational designs, a range of variables are measured. These designs are also called correlation studies, because correlation data are most often used in analysis. Since correlation does not imply causation, such studies simply identify co-movements of variables. Correlational designs are helpful in identifying the relation of one variable to another, and seeing the frequency of co-occurrence in two natural groups (See correlation and dependence). The second type is comparative research. These designs compare two or more groups on one or more variable, such as the effect of gender on grades. The third type of non-experimental research is a longitudinal design. A longitudinal design examines variables such as performance exhibited by a group or groups over time. See Longitudinal study.
Examples of flexible research designs
See also: Case study
Famous case studies are for example the descriptions about the patients of Freud, who were thoroughly analysed and described.
Bell (1999) states “a case study approach is particularly appropriate for individual researchers because it gives an opportunity for one aspect of a problem to be studied in some depth within a limited time scale”.
See also: Ethnography
This type of research is involved with a group, organization, culture, or community. Normally the researcher shares a lot of time with the group.
Grounded theory study
Grounded theory research is a systematic research process that works to develop "a process, and action or an interaction about a substantive topic".
- ^Muaz, Jalil Mohammad (2013), Practical Guidelines for conducting research. Summarizing good research practice in line with the DCED Standard
- ^Robson, C. (1993). Real-world research: A resource for social scientists and practitioner – researchers. Malden: Blackwell Publishing.
- ^Adèr, H. J., Mellenbergh, G. J., & Hand, D. J. (2008). Advising on research methods: a consultant's companion. Huizen: Johannes van Kessel Publishing. ISBN 978-90-79418-01-5
- ^Bell, J. (1999). Doing your research project. Buckingham: OUP.
- ^Creswell, J.W. (2012). Educational research: Planning, conducting, and evaluating quantitative and qualitative research. Upper Saddle River, NJ: Prentice Hall.
This is an introduction to doing research, particularly original research. Please feel free to edit this guide, to add methods, methodologies that you use, or to add questions, requests or comments on the talk page.
You will most likely start doing research because you have a particular interest in a particular field. You might want to find out more about, for example, the links between poverty and crime; how to provide for multiculturalism in the classroom; or what the extent and effects of pollution in your area are.
To find out more about this interest of yours, you must identify a "path" that you will take in order to undertake this research. You will most likely not know what this path is, but you might have some idea of where to start - eg. from your theoretical/disciplinary standpoint, or from something that you have read that has made you think, even if you disagree with it. You may find, in thinking about your area further, that there is something which is not working, or which is unknown, or perhaps which is hypothesised, but that needs to be tested. This is the context for your research - your research problem. The next thing that you need to do is to turn that problem into a question or a statement - which you will use to address this problem.
Your research question (or questions) should be your tool(s) for addressing the issue that you have identified as being of interest to you. The way you ask the question is vital to determining what kind of research you will conduct. For example, if you are interested in the second example above - multiculturalism in the classroom - you could ask a number of questions about this, all of which will guide you in a specific direction. Examples of questions to address this context/problem might include:
- "Why are some schools managing to integrate students from different backgrounds better than others?"
- "How are teachers in [X] coping with the increasing numbers of students from [X]?"
- "What could I do to improve the intercultural awareness of the students within my class?"
- "What is the impact of multiculturalism on classroom ethos?
- "What are the strengths and weaknesses of multiculturalism in the classroom?
Each of these questions has a particular slant (possibly even a philosophy), both in what it is targeting and how it is phrased. They will also inevitably spawn a number of other questions, or sub-questions. They also may need to be refined, or clarified (such as, in the second question, by asking "What measures are teachers taking to cope with ...?". This is a continual process that you will have to think about constantly throughout your research - possibly even after your data collection and analysis. Things to bear in mind in forming questions to ask is to be realistic in what you can answer (with the time/resources you have available), and also in how many questions you are answering (better to have one or two well-focussed questions, than five vague ones).
Research cannot exist in a vacuum. In order to be scientific and rigorous, your research must itself be based within the context of "the literature" (ie books, journals, newspaper articles). Literature here can be taken broadly - it is perfectly valid, for example, to cite television programs as contributing to the context of your area of inquiry. Your research should show that you have read around both your subject and the methodologies that you have chosen - your questions, methodologies and methods will also largely be shaped or influenced by what you have read.
Methodologies and methods
There are a wide array of research methodologies and methods, and, while there are some distinctions amongst these, there can also be significant overlap or multiple methods/methodologies used in a single research design. Research methodologies can take the form of experiment, case study, and/or survey, can be either, or a mixture of, qualitative (based on words and meanings) or quantitative (based on statistics and their meanings), and can incorporate a variety of methods to generate data (eg. observations, questionnaires), as well as varieties of ways of analysing this data. The following are some common ways of designing a methodology that answers your research question(s), and methods of generating data.
- Experiment: An experiment-based methodology is where, simply speaking, a stimulus is applied (eg. a new system of teaching science to primary school students) and its response is measured (eg. by analysing exam results). Such a methodology is most often linked with a quantitative (ie statistical) approach, but this is not necessarily the case. To maximise the validity of such studies, there is usually some element of controlling of/for variables (such as by having a group of students who are taught differently to normal, and another group who are taught the same as normal). It can be linked with methods such as observation, interview etc.
- Survey: A survey is a study of a phenomenon over/within a geographic region. This could involve, say, a survey of the crime rates of every major city in a certain country (where "major city" needs to be defined), or a survey of a sample of bloggers' political motivations (where this sample needs to be defined).
- Case study: A case study, as the name implies, is a study of a specific "case", or group of "cases" - a "case" being an individual person, an organisation, a school etc. Some research will focus on one single case and attempt to generate "rich" data (ie revealing as much complexity as possible); and some research will focus on a number of cases, either which are significantly different from one another, or which are similar, or which are clustered or spread in a geographic/socio-political spread. Focussing on a number of cases can approach a survey design (or mini-survey) - or sometimes large-scale surveys can be used in order to identify specific cases which might be of interest to the researcher.
Overall, research design is a complicated, and personal, thing. There is no research which is implemented from another design "off-the-shelf". Of course your research will probably echo much research done before - and this is a good thing - but, in order to be individual, interesting and useful, it needs to be continually grounded in the research questions that you have outlined - as well as appropriate to the subject/context/environment/population that you are studying. In order to address a complex question (and all research questions are complex), you will need to identify what methodologies and methods - or, more likely, combinations of methodologies and methods - are most likely to address your particular question to your satisfaction.
- Sampling - Are the people you have chosen to participate in your research (or who have themselves chosen to participate in your research), representative of the population? In other words, if you can draw conclusions from your particular study, will it be useful or applicable to other people? Does it matter to you if your sample is unrepresentative?
- Ethics - Are you potentially harming someone through your research? A central maxim of human-subject research (ie that involving people) is "do no harm".
- Validity - How might your design be flawed, or your conclusions wrong?
Some examples of research methods:
- Questionnaires: Sometimes a questionnaire can contain a number of questions with a number of options to choose from (ie where you have to "tick a box, or number of boxes"). Other questionnaires may be questions with space in which to write more free-form or detailed answers. Some questionnaires will have a mixture of both types of questions. Both types of questions can have their strengths and weaknesses.
- Observations: Observation is paying close attention to an environment, its context, and its social dynamics. It can be systematic (where the researcher will be recording, for example, how many times a person scratches their head), or more free-form where the researcher watches everything and records as much detail as they can or that they feel is appropriate. This latter type is a form of "participant observation", and which is often associated with anthropological or ethnographic research.
- Interviews: Interviews can be between one person and another, or in a group setting. They can be "structured" (where the interviewer will ask a predetermined set of questions), "semi-structured" (where the interviewer will ask a number of questions based on an outline of topics to be covered), or "unstructured" (where the interviewer will ask questions based on whatever emerges during the interview itself - or, often, will not seem to ask questions, but rather facilitate or participate in a conversation).
- Eliciting: Eliciting is a way of getting people to talk about something, based on a prompt, such as a photograph, or piece of music. For example, in research done with young children, an interview might be intimidating, but a photograph (for example) gives the child or children something to talk about, while giving the researcher an opportunity to observe reactions to the photograph.
Return to the Research questions section above, and try to see what methodologies and methods you would use to address each of them. Why do you think your selection of methodologies and methods works better than other possible options? What potential issues do you see arising from your choices?
Once you have collected your data (e.g. filled-in questionnaires, interviews recorded and transcribed), you must now do something with it! What good is your data to anyone else if it is not interpreted (except, of course, other researchers)?
Writing up your research into a report, paper, essay or thesis. See for more at Help:Resource types