Explain a census and the impracticality of conducting one in social science research.

The Research Process

This course is aimed at the person who will APPLY data to daily communications problems rather than the statistician.

Many professionals who use statistics such as ratings to make important decisions do NOT understand the limitations of the procedures used to derive the statistics. Remember that a statistic is JUST a number until we understand the data to which it is applied.

The history of media research is full of practical studies which have had theoretical implications.

The hypodermic theory was upset by a study of election polls by Katz and Kahn. They found that the media affect a few opinion leaders who in turn affect small groups, etc.

A contemporary issue is the movement of many mass media toward class media. This trend has important implications for people in programming and marketing.

The trend today is toward inter-disciplinary studies which can help bring new points of view to bear on a question.

The most important division in the research community falls between proprietary researchers and those in academic research (also called public research). Research always begins with a question. The best questions are frequently those close to home. Look around before going off in left field.

There are 4 ways of answering a research question.

1. tenacity: holding with tradition

2. intuition: generalizes from internal data, usually inadequate

3. authority: relinquishes responsibility to others

4. scientific method: approaches an issue in a series of small steps, each of which is tested objectively. This is powerful and in the long run self correcting, but we can never PROVE things with this method, we only strongly suggest.

Scientific research should be public. The Reagan administration was criticized for trying to constrain publication of pure research findings in sensitive areas during the Cold War. This hurt America more than it did our enemies because our one strength has been in our open exchange of ideas.

There are 7 steps to the scientific method:

Are heavily inter-related

In tackling research problems there are three approaches which can be arranged in a hierarchy based upon the researcher's control of the environment.

l. Descriptive: no attempt made to advance theories or control variables. Data are presented as they lie. Looked down upon by journals, but vital to the real world professional. Do not try to impose an advanced approach upon a good descriptive study.

2. Empirical: from observation the researcher speculates upon relationships between variables which he or she can not control. Can be effective and in many social settings it is the best we can do. Has the advantage of the natural setting.

3. Experimental: most powerful and respected, but usually must be conducted in an artificial environment. The experimenter manipulates one or more variables and observes reaction in others.

In experimental studies we talk of the independent and dependent variables. In empirical studies the dependent variable is called the criterion variable.

We attempt to control all variables except the ones we want to observe. This is hard and leads to intervening variables--those which affect outcome but of which we are not aware.

There is a current trend toward multi-variate studies because they allow you to determine which of many variables are most important to the behavior or setting being studied.

The qualitative vs. quantitative dichotomy frequently discussed in research textbooks has little value. Journals like quantitative.

Measurement of Relationships--Some terms to think about:

Research Ethics

The most important lesson any researcher can learn is that the data are! Unexpected results can be the result of a poor design, a biased sample, or they just MAY be the key to a major breakthrough in our understanding of communications. Fudging your data to conform to a preconceived hypothesis is not only dishonest, but may cause you to miss an important relationship which may be far more important than any of the ones you set out to investigate.

Humans being human, it is very difficult to be totally objective about a research project you have conceived and developed. For this reason it is critical to craft your design to minimize the impact of your preconceptions. In medical research, the gold standard is the double blind experiment. In the social sciences we are seldom this exacting, but it is usually possible to distance yourself from any qualitative evaluation required by your design. We will cover this in more detail as we examine different designs.

Return to TVR 400 Home Page