Experiments
Experiments are aimed to study casual links. The simplest one is about if there is a link between two variables while some complex ones may consider the size of the change and the relative importance of variables (Saunders, Lewis and Thornhill, 2009). Therefore experiments are designed to answer ‘how’ and ‘why’ questions and all the relevant variables are controlled by experimenter rather than simply observed in natural setting. Experimenters provide a valid and clear test of the research hypotheses by managing for all possible random variables. This aims to test the effects of independent variables. Hence, the main strengthens of experiments is that strong conclusions can be figured out about cause and effect relationships. By contrast, the key weakness is that they usually present unvanquishable practical and ethical limitation to the control of people (Williamson, 2009). The experiments often proceed in laboratories instead of the field. Thus, people have greater control on the research control like sample selection or variables selection. This raises the internal validity of the experiment while the external validity seems to be harder to establish. In summary, experiments are high in reliability and internal validity, but low in external validity.
The Hawthorne studies is a kind of psychological experimenter effect which means that the observed one changes the tendency of behaviour when one realized that he or she is being observed. The study is designed to see if worker would become more productive in higher or lower level of light. As a result, the workers’ productivity enhanced after the study started. The results obtained in this experiment are similar, so it is high in reliability. The setting of Hawthorne studies has a lot of human factors and cannot be generalized from naturally occurring conditions. However, the extraneous variables are controlled. In this way, it is low in ecological validity and high in internal validity.
Surveys
Surveys are one of the deductive approaches and popular in business and management researchers (Saunders, Lewis and Thornhill, 2009). People tend to use surveys not only because the large amount of population but also the low cost. There is always no need to do a survey among the total population. Samples must be the representative of the total population and should be typical of the population, then, it is possible to generalize to the total population. This reveals that surveys are high in population validity. Surveys require a sample of people to do the fixed questionnaires and comparison could be made. Thus surveys are often collected in a quick and economical way. But sometimes the answers of the survey may be superficial (Williamson, 2009). To sum up, surveys are usually highlight reliability and generalization and little emphasis on the setting or the description of explanatory variables.
Surveys are common in the daily life and always used as a method to collect data. When doing my undergraduate dissertation, I considered surveys as my main approach to collect data. The survey is about the cosmetic brand strategy. The sample is fixed in a particular region. After finished all the surveys, the key step is to compare the results and make a conclusion. It is then clear from the example that surveys are high in reliability and population validity but a little low in internal validity.
Ethnographies
Ethnography is an inductive method. The aim of ethnographies is to describe and explain the social world the research followed. This is a time consuming approach and happens over an extended time. The main advantage of ethnography is its focus on naturalism, but the drawbacks are time consuming and low representativeness and reliability. The stress on the setting and description of explanatory variables is strong (Williamson, 2009). Therefore, ethnography is high in ecological validity and low in internal validity and reliability.
Timothy M. Waring, a human ecologist, believed that learning a culture the way a child does, by living, working, speaking, and participate it in. The longer the better (Waring, 2011). He spent a year and a half in the Palani Hills and focused on learning the rural agricultural villages, doing some farm work, speaking native language and try to see the world like local farmers. After such a long period observing and studying, Waring learned a lot.
Other Key Factors When Choosing a Research Design
Time Horizon
When one is planning a research, time is the key factor which cannot be neglected. Whether choosing a short period research or a long time one, it depends on the research questions. The snapshot time horizon is called cross-sectional while the long period like ‘diary’ we call longitudinal (Saunders, Lewis and Thornhill, 2009).
The cross-sectional studies are suitable for a particular phenomenon in a particular time. Most academic research programs are time constrained. So surveys always use cross-sectional studies as it always finishes in a short time. While sometimes researchers need to observe people or event in a long period. This is called longitudinal studies and the best example of it is ethnography. But one thing should be sure in longitudinal studies is that is there any change during such a long time.
The Ability of the Researcher
People are the essential factor among all the factors during the research. The ability of a researcher contains the knowledge, the skill and technical background. Researcher as the creator should know and be acquainted with all types of researcher designs. No matter which research design one choose, researchers have the ability to evaluate it. This demands researchers acquire sufficient knowledge in academic field and good scientific research ability. It would be ridiculous if researchers have no strict scientific ability to handle it.
Apart from the academic skill researchers should have, the ability of communication is also very important. When choosing a research design, researchers not only need to search materials but also need to communicate with partners or their guiders. It dramatically improves the efficiency of work. These people will share the experience and give advices to researchers, all these useful information helps researchers to decide the type of research design.
Availability of the Money
The cost of the research is also an important element to consider before choosing a research design. Researchers should be clear how much fees he or she own and distribute reasonably into every steps. Different research methods have different cost. It varies on what approach one chooses. For example, when doing a survey on a sample, researchers usually choose to collect a sample of the population instead of trying to survey everyone, as studies cost more money the larger the sample size gets. The smaller the sample size is, the less money it will cost. Therefore, the research design also depends on the availability of money especially when the money is limited.
Conclusion
To sum up, these three research designs have their own characters. In experiments, the cause and effect relationships are readily determined, but have some practical limitation in social research. Experiments are high in internal validity and low in external validity. It is also high in reliability. Refer to surveys, it is well known among people and is the common methods for using. Not only as it gathers information quickly, but also it is cheaper than other methods. One backward of surveys is that some answers may lack of profound. It has high reliability and population validity. However, ecological validity and internal validity is a bit low, extend to medium to low. The last but not least, ethnographies have a strong emphasis on naturalism. But it is low in reliability and population validity. Moreover, the internal validity is arguably low. By contrast, the ecological validity is high.
After contrasting the three research methods, it comes to the key factors should be considered when choosing research design. Time horizon, money and ability of the researcher are all regarded as the necessary factors to consider. Time horizon controls the start dates and ends date and makes researchers manage every step in a reasonable time. Then researchers have to put the money into account. It directly decides the scale of the research and how much employees should hire. So doing a cost budget before choosing a research design is vital. The third element is about the ability of researcher. There is no need to point out the importance of a researcher. Researchers tend to be a guider or a leader among the whole progress. They may be assessed as how much knowledge do they acquire, what skills have they owned, or what technical background they have. It is the key factor among others and has a big influence on research design. All these factors help researchers to figure out the perfect research design and make the progress going successfully.
Reference
-
Gill, J. and Johnson, P. (2010) Research Methods for Managers (Fourth Edition), Chapter 9
- Mitchell, M. and Jolley, J. (2001). Research Design Explained (4th Ed). New York: Harcourt.
- Saunders, M., Lewis, P. and Thornhill, A. 2009. Research methods for business students. Essex: Pearson Education Limited.
- Thietart et al. R. 2001. Doing management research. London: SAGE Publications Ltd.
- Waring, T. 2011. Methods [online]. Available from: http://timwaring.wordpress.com/cv/methods/.
- Williamson, G. 2009. Research Styles [online]. Available from: http://www.speech-therapy-information-and-resources.com/research-styles.html