Research 201: A Basic Introduction
© Brian Brown 1999-2000. All rights reserved.
Different Types of
Research can be conducted in a number of different ways for many different purposes. Most research conducted today is in the corporate sector, and the reasons for this should be obvious.
Developed since the 1930's, survey research seeks to identify what large numbers of people (mass) think or feel about certain things. It is used extensively in politics and marketing (such as TV advertising).
Examples of survey research are
An amazing fact about survey research is that the amount of error (expressed as plus and minus a certain percentage) is determined by the sample size (the number of people surveyed). Most opinion polls use a sample size of around 1500, which has a margin of error of ± 3%. Using a larger sample size than 1500 gives a slight reduction in the error margin, and using a smaller sample size than 1500 significantly increases the error margin.
Surveys are often considered biased because
Increasingly we find the Internet being used to conduct survey research, with the use of opinion polls and questionnaires. This is due to the ease of creating on-line questionnaires and the power of analysing the data in real-time by the use of powerful database servers.
A focus group gathers in-depth information by interviewing six to twelve experts in an informal discussion that lasts one to two hours. An experienced interviewer gathers opinions of the group.
Benefits of Focus Groups
Problems Using Focus Groups
Typical Uses of Focus Groups
A case study looks at existing information that is readily available and attempts to draw conclusions from this information. Information has been collected and stored from a wide variety of sources about many issues. Much of this information is readily available in the public domain.
The researcher begins with a number of research questions. For instance, a researcher might want to look at the impact of information technology in South Africa since the overthrow of white rule. The researcher will identify information in the public domain, such as tele-density, number of Internet hosts, and other data. From this information, the researcher will draw qualitative conclusions.
Case studies allow existing information to be reused in new ways that the original collectors of the data did not envisage.
The Delphi method surveys the opinions of "expert panels". The research is conducted in three rounds, where the information is gathered, refined and then feed back to the expert participants. The feedback stage allows issues to be sharpened and helps to highlight the major issues involved.
Round one involves an in-house panel that seeks to identify the research questions for the experts. It also conducts a pre-test of the survey.
The second round involves an examination of the issues by the expert panel. At this stage, additional items are often suggested that may have been overlooked. The panel responses and additional items are combined to the original questionnaire.
In the third round, the panel make judgements on the items. This sometimes takes the form of a multiple-choice questionnaire.
Content analysis is often used in quantitative research to study trends or occurrences of information. During World War II, the allies monitored the number and types of songs played on European radio stations. By comparing the music played on German radio to that of other radio stations in occupied territory, the allies were able to measure the changes in troop concentration on the continent.
Other early uses of content analysis were attempts to verify authorship of documents. Knowing that certain authors chose certain words, many scholars attempted to show that some documents had undergone a process or redaction or rewriting by subsequent editors.
Content analysis is heavily involved in the use of television. A common exercise is to use content analysis to measure the incidences of violence on television or in print.
There are three main characteristics associated with content analysis
How is content analysis done? Let us consider an exercise. Our research hypothesis is that "More than 75% of leading news items are concerned with violence". This is based on the media assumption that "if it bleeds, it leads". Firstly, we have to define what we mean by violence. Since content analysis of the mass media is a well-established field, there are many sites on the Internet that can assist in providing useful definitions. Let us consider that violence in this research means bodily violence inflicted by one person on another, resulting in injury, hospitalisation or death.
Having decided on what violence means, we then monitor the lead news items over a number of weeks. An adequate sample size must be chosen to eliminate bias, as it is possible that a particular news item could be repeated over several days (for instance, a war could break out). We might decide to select three days a week at random over a period of three to four weeks.
Coders are responsible for coding each story into a violent/non-violent category. There is a form for each day that the coder fills in. These forms are collected at the end of the study and then the results tabulated accordingly.
The coding form shown below is a more comprehensive one for analysing each item of the evening news broadcast. Coders use this form to mark the category of the particular news item under consideration.