Once the why and what of research has been established it is time to tackle the important question of ‘How to do research?’ This blog will try to explain the various steps involved in the research process.
The starting point has to be identifying the need for research. Good research involves committing a lot of resources- tangible & intangible, and therefore it is imperative to have a clear picture of the need for research. As the research process and inferences are guided by the established need for research, it has a direct bearing on the outcome of the entire research. It is also important to identify situations where research is not needed, like if the data that is to be collected is already available or if the costs of the research outweigh the corresponding benefits. In such cases a conscious decision has to be made whether to pursue the research or not. Also research does not help with problems that require immediate solution.
The next step is to define the problem or formulating the research problem statement. It is the most important step as an incorrect definition of the problem will render the entire research pointless. So it is important to identify the correct problems and articulate them in the correct statements. Problems can be identified to stem form gaps between what is supposed to happen and what did happen & gaps between what did happen and what could happen. The research problem has to be specific, well defined and summarized in one line.
The third step is establishing the objectives of research. Objectives of research are the steps taken to carry out the research. Research objectives although determined by the problem definition, are set so that when achieved, they provide the necessary information for the solution. Research objectives can be used to identify & assess either the critical (low performing) processes or the ideal (high performing tasks).
Focused Group Discussion:
It is a form of qualitative research where a group of people are asked about their opinions and attitudes about a particular product. The method can be used for various intentions. Marketers use it a feedback tool to gather customer’s experiences about a particular product. Social services use the method in policy formulation and identifying solutions to existing problems. The method is especially useful to understand the a given problem and its impact on the system holistically from multiple perspectives.
Types of FGDs:
- Two-way focus group
- Dual moderator focus group
- Dueling moderator focus group
- Respondent moderator focus group
- Client participant focus groups
- Mini focus groups
- Teleconference focus groups
Some of the disadvantages of FGDs are – experimenter’s bias, groupthink, social desirability bias. Unless the discussions are repeated over multiple iterations, the results can cause severe issues in terms of external validation.
Proportionate Hazard Analysis:
These models are part of survival models in statistics. Survival models are used to predict the time elapsed between the occurrences of successive events such as a biological death or a mechanical failure. These models are extensively used in biomedical & health services research. The concept bears close resemblance to reliability theory in engineering or event analysis in political sciences.
The objective of such analysis is data in the form of times from a well defined time of origin and to an end point. It is used in research to identify the problem more clearly and to understand the various processes preceding and succeeding the problem and their impact on the problem itself.
It is one of the common forms of experimenting where each individual trial is compared with a known control so as to ascertain the exact impact of the change. Here the control has to be specified and the trial should have only one change as compared to the control. It two changes are to be tested the two separate experiments have to designed. This eliminated ambiguity and helps in establishing the exact impact of each individual change of the process. A separate experiment has to be designed to identify the combined effect of the two changes on the system as compared to the control. The main objective is not to cascade one trial over the previous ones as this will not help in indentifying the individual impact of each trial.