Sociology as Science
January 11, 2015
Social research is a process for producing new knowledge about the social world in a structured, organized, and systematic way (Neuman, 1994:2).
Why is social science (sociology) science? Is sociology simply a pseudo-science? After all, its ability to predict the future is questionable! Isn't it? What is science? In mathematics, 2 + 2 always = 4. Sociology often cannot make precise predictions.
In response, one might argue that just because the subject matter of sociology is more difficult to study than the subjects pursued in other sciences, it does not mean that the scientific method is inappropriate for the social sciences. The subject matter of sociology experiences continuous change. This fact alone renders efforts at prediction difficult. Problems relating to prediction can be found in the biological science as well. One should note the problems encountered as biologists try to track the AIDS virus. It too continually mutates.
Sociology is a science every bit as much as biology or chemistry. Social sciences, like natural and biological sciences, use a vigorous methodology. This means that a social scientist clearly states the problems he or she is interested in and clearly spells out how he or she arrives at their conclusions. Generally, social scientists ground the procedure in a body of existing literature. This is precisely how other sciences function.
The scientific method of understanding society is relatively new in the grand course of human history. It arose during the Enlightenment period in Europe during the eighteenth and nineteenth centuries. Before exploring scientific sociology, let's begin with a brief discussion of other sources of social knowledge about society. I do this for two reasons:
In order to understand where we are, it is sometimes helpful to understand where we have come from and where we are going (with the lectures to follow). That is why we study history!
Further, a good way to determine the worth (or lack of worth) of anything social is within a comparative context that offers alternatives.
Often, we get our knowledge from significant others like parents, teachers, books, or political leaders. When one accepts something as true because someone in authority says it is true, then they are relying on authority. It is a quick, simple, and inexpensive way to gain information (Neuman, 1994:2-3).
The problem associated with relying on authorities is that overestimating the expertise of someone or some publication is possible. An expert in one area might ...
- try to use his or her expertise in an area where the authority has little if any knowledge. Neuman (1994:3) reminds us that "experts" used to measure intelligence by counting the number of bumps on the skull.
- An over reliance on authority may also produce problems in a democratic society. Allowing authorities to wield too much authority can be dangerous! Over reliance on authority might lead to dictatorship.
Neuman (1994:3) contends that tradition is a special case of authority, the authority of the past. "It has always been done that way." One problem with relying on tradition as a source of information is that conditions change. People can cling to past traditions without understanding why something was true in the past (e.g., A shot of whiskey cures a cold). Tradition can also be based on simple prejudices that people pass down from one generation to the next. Even if traditional knowledge was once true, it can become distorted over time. (E.g., The best way to plow a field is with a mule-drawn plow, or one should always plant by the full moon.)
Common sense is the knowledge people gain about the world through their everyday experience. It works sometimes. In fact, sociology might require that one use a little common sense when engaging in research projects. On the other hand, one still has to remember that common sense is not truth in any objective sense. It is only a shared social idea that people find comfortable and safe.
Example: Simple Dichotomies
The seemingly persistent tendency for human beings to think in terms of simple dichotomies to understand society perplexed Fernando Henrique Cardoso. To emphasize the simplicity of such thinking he used the metaphor of two space travelers encountering earth for the first time. The space travelers express shock at the simplicity of the earthlings. They might say, "the brain of these beings appears to limit their images and thoughts to binary opposites" (Cardoso, 1977).
Example: Who is Rich, Who is Poor?
Asian-Americans have the highest per capita median income in the United States while Native-Americas have to lowest. This contradicts the usually accepted notion that Blacks and Whites define the top and bottom of American society.
1. Problems with Common Sense
a. Our Experience is Limited
We cannot possibly know everything everywhere.
b. Our Interpretation of Experience is Biased
Our minds play tricks on us. We are likely to see what we want to see. We are likely to look for easy explanations and we are likely to accept ideas of people that are attractive to us. Sociologists have dubbed this tendency the "halo effect."
Example: What is Suicide?
When is a death suicide? If someone attempts to fake a suicide but actually succeeds in killing themselves, is their death suicide or accidental death?
Example: Suicide in Religious Communities
Some religious communities show a low suicide rate. Does this mean that people in these communities kill themselves at a lower rate? In some religions suicide is a mortal sin. Perhaps religious communities attempt to cover up suicides more than non religious communities. One explanation might be that suicide in religious communities would have more serious social impact on the survivors than it would in non religious communities.
Example: The Problem of "Illegal Aliens"
Common sense tells us that undocumented workers take jobs from Americans and that, in general, they create a burden for the U.S. taxpayers. Facts, however, show us that undocumented workers add more to the United States economy than they cost. Further, they tend to take jobs that most Americans don't want.
Example: Buy American! What Does This Mean?
As Americans struggled with the global economy in the 1980s, many advocated buying American products from American companies. Common sense told us that buying American would put Americans to work and make the U.S. economy stronger. Unfortunately, distinguishing between global and domestic economy became highly problematic in the 1990s as the domestic and international economies became more interconnected (See Keohane and Nye, 1977; Reich, 1991).
General Motors is an American company, but look at the international involvement the creation of a General Motors product like the LeMans (from Reich, 1991).
Of the $20,000 paid to GM, about $6,000 goes to South Korea for routine labor and assembly operations, $3,500 to Japan for advanced components (engines, transaxles, and electronics), $1,500 to West Germany for styling and design engineering, $800 to Taiwan, Singapore, and Japan for small components, $500 to Britain for advertising and marketing services, and about $100 to Ireland and Barbados for data processing. The rest -- less than $8000 -- goes to strategists in Detroit, lawyers and bankers in New York, lobbyists in Washington, insurance and health care workers all over the country, and General Motors shareholders -- most of who live in the United States, but many who are foreign nationals (Reich, 1991:113).
This one is obvious. Have you ever heard Arnold Schwarzenegger say "Hasta la vista baby" for George Bush? The TV is notorious for distorting reality about crime, romance, etc. The news also can distort truth whether intentionally or otherwise (to meet deadlines, etc.) (Neuman, 1994:4).
The scientific method is a systematic, organized series of steps that ensures maximum objectivity and consistency in researching a problem (Schaefer and Lamm, 1992:35). The following are some components of the scientific method.
Don't take assumptions for granted. Don't rely on common sense. Don't rely on traditional authority figures.
Evidence should be observable because other Sociologists might want to perform the same study in order to verify or refute findings.
1. Social Facts
Henslin (1999:16) notes that Durkheim stressed social facts. He calls them "patterns of behavior that characterize a social group." Appelbaum & Chambliss (1997:12) defines social facts as "qualities of groups that are external to individual members yet constrain their thinking and behavior." For example, one may display a particular behavior when with friends, but feel constrained to act differently when in a more formal setting. The effect of a social group on individual behavior is a social fact.
Any study of society should specify the methods the researcher used to obtain his or her information, the setting (where the researcher conducted the study), and the population (whom they studied). This is done so that other social scientists may test your findings. Social scientists are cautious in accepting the findings of other. Studies are often replicated to verify findings of initial studies.
A theory is a set of ideas [generalizations] supported by facts. Theories try to make sense out of those facts. Social scientists seldom accept theories as laws. Often they are not considered totally true. Furthermore, the subjects they attempt to explain (i.e., people and social institutions) are variable. Gergen (1982:12) in D'Andrade (p 27) states:
"It may be ventured that with all its attempts to emulate natural science inquiry, the past century of sociobehavioral research and theory has failed to yield a principle as reliable as Archimedes principle of hydrostatics or Galileo's Law of uniformly accelerated motion."
Because theories are general ideas, social scientists do not test them directly. A hypothesis is a speculative (or tentative) statement that predicts the relationship between two or more variables. It is, in essence, an educated guess. It specifies what the researcher expects to find. To be considered meaningful, a hypothesis must be testable; that is, capable of being evaluated (Schaefer & Lamm, 1992: 38).
The mean, or average, is a number calculated by adding a series of values and then dividing by the number of values. For example, Eleven students who completed the first test had scores of 48, 57, 64, 68, 68, 70, 78, 84, 90, 92, and 95. In order to determine the mean, add the eleven scores and then divide by the number of scores (11). The mean is 74.
The median is the midpoint or number that divides a series of values (which are ranked in ascending or descending order). Eleven students who completed the first test had scores of 48, 57, 64, 68, 68, 70, 78, 84, 90, 92, and 95. The median for this group grade of students is 70.
A percentage is a portion based on 100. Use of rates (and percentages) allow one to compare populations of different sizes.
Example: Comparing Populations of Different Sizes
If we are comparing contributors to a town's Baptist and Roman Catholic churches, the absolute numbers of contributors could be misleading if there were many more Baptists than Catholics living in the town. With percentages, we can obtain a more meaningful comparison, showing the proportion of persons in each group who contribute to their respective churches (Schaefer & Lamm, 1992: 36).
In a sociological sense, control means that you neutralize all social characteristics (variables) except that which is under consideration. This is different from a control group. A control group is something associated with an experiment. If one is testing, say a new drug, one would get two similar populations. The new drug would be given to one group (an experimental group) and withheld from the other group (a control group). Any difference between the experimental group and the control group is probably due to the intervention (e.g., the new drug in this example).
The target population refers to everyone in a group that is studied. For example, if one wants to know how people will vote in an election, the target population is everyone who is eligible to vote. How can a researcher study a population as large as that of the United States? The answer is that one cannot study entire populations. Large populations are simply too big. The researcher, therefore, needs to look at a small subset of the population. We call this subset a sample. The trick is to make sure that your sample closely parallels the characteristics of the larger population.
1. Random Sample
Henslin (1999:126) contends that a random sample is one in which everyone in a population has the same chance of being included in a study. A random sample is necessary if one is going to attempt to generalize the findings in a study to the larger population.
Generalizability refers to a condition where a social scientist is able to apply their findings (drawn from a sample) to the larger population.
A hypothesis poses a relationship between two or more aspects of social relationships. These aspects are called variables. A variable is a measurable trait or characteristic that is subject to change under different conditions. Income, gender, occupation, and religion are variables. Variables may be independent or dependent.
1. Independent Variables
Independent variables in a hypothesis are those that influence or cause changes in another variable. In other words, an independent variable is something that is chosen by the researcher to cause a change in another variable.
2. Dependent Variables
The dependent variables are those variables are believed to be influenced by the independent variable (Schaefer & Lamm, 1992:38).
Example: Independent and Dependent Variables
Higher levels of education produce greater earnings. Education is the independent variable (it causes the change in income levels). Income level is the dependent variable. The income an individual earns "depends" or is determined by the influence of education.
One of the most common research mistakes is to assume that a high correlation between two variables means that one variable (independent) causes some change in another variable (dependent).
A correlation exists when a change in one variable coincides with a change in another variable. The two variables are regularly associated with one another. However, the mere fact that associations exist, does not necessarily mean that a change in one variable causes a change in another variable. Correlations are an indication that causality may be present. They do not necessarily indicate causation (Schaefer & Lamm, 1992: 38).
A spurious correlation is one where the apparent correlation between two variables is actually caused by a third variable (Henslin, 1999:130)
One of the most common research mistakes is to assume that a high correlation between two variables proves that there is a causal link between them. In other words, people assume if two variables are related, then obviously one causes the other.
Causality is rather difficult to demonstrate. How can one tell whether a change in one variable is "causing" a change in another variable? There are three requirements that must exist before one can begin to think about whether there is a cause and effect relationship.
1. Temporal order:
The independent variable has to occur before the dependent variable.
2. Association (or correlation):
A change in one variable is associated with a change in the other variable.
3. Elimination of plausible alternatives:
The researcher has to ensure that the association between the two variables is not caused by a third variable (e.g., there are no spurious correlations). In order to show that one variable cause a change in another variable the scientist has to control for other factors that might be influencing the relationship.
4. Does it make sense?
Finally, there is also an implicit fourth condition. The causal relationship has to make sense or fit within a theoretical framework (Henslin, 1999:131).
Validity exists when concepts and their measurement accurately represent what they claim to represent while reliability is the extent to which findings are consistent with different studies of the same thing or with the same study over time.
Weber suggested that sociology needs several methods of investigation. The following material provides various benefits and problems associated with four methods of gathering data.
- Case studies (or field studies) explore social life in its natural setting, observing and interviewing people where they live, work, and play (Kendall, 1998:25).
- Its advantages are that the researcher can study individuals in their natural setting (e.g., at home, at work, playing, etc.). Case studies provided volumes of information such that at the end of the study the researcher has a thorough understanding of the individuals involved in the study.
- Drawbacks to the case study include the fact that social scientists cannot usually investigate many cases because of time constraints. Another problem with the case study is that the results may not be generalizable to the population at large.
- The researcher asks questions of the cases face to face or in a questionnaire.
- The advantages are that data collection is more systematic (you ask the same questions of every case).
- Because it is systematic and generally more condensed, the researcher can investigate more cases. Survey research can, in fact, be applied to several thousand (or million) cases. The U.S. Census begins as a survey of the population.
- Findings may be generalizable to larger populations.
- When relying on a survey questionnaire, much information is lost. Facial expressions are not recorded. Environmental considerations are missed.
- Furthermore, information can be lost because the interviewer failed to ask the right question.
Kendall (1998:26) describes an experiment as a "carefully designed situation (often taking place in a laboratory) in which the researcher studies the impact of certain factors on subjects' attitudes or behaviors."
- The experiment offers a high degree of exactness because one can control everything in a laboratory setting.
- Variables can be precisely studied. Natural science uses this approach most often. So does psychology.
- It is easier to determine cause and effect relationships.
- One disadvantage with the experiment in studying social phenomena is that the environment is contrived. People do not normally carry out their lives in a laboratory setting.
- Ethical issues may also arise when performing experiments on people. The Nazi death-camp experiments represent extreme instances of ethical violation. Even in ordinary university type experiments deception and misinformation are often employed. Many consider these ethical violations.
- Existing data includes government records (census), personal documents, or mass communication (published books, the news, movies).
- The Statistical Abstract of the United States is an excellent source of existing data.
- The advantages are that data are generally easy to obtain. They already exist and can be found in most university libraries.
- Much existing data are also standardized. Standardization makes it easier to compare one set of data with another.
- One problem associated with existing data is that the researcher must use the format provided. For example, a researcher studying poverty would be frustrated with the census before 1970 because there was no poverty rate in 1960 and before.
The scientific perspective might cause one to look for cause and effect type relationships. Researchers may assume relationships are cause and effect where, in fact, many actions undertaken by individuals, groups, etc. involve choice. Further, while one expects cause and effect to travel in one direction, it may actually travel in the opposite direction. Furthermore, what may appear to be a cause and effect relationship between two variables may be driven by a third variable.
Example: Science as a Bias
Science itself may play a role in how one interprets a given social phenomenon. Furthermore, it may influence the solutions for social problems.
is the Hawthorne effect?
People who hold positions of power within universities, private enterprise, and the government have the power to decide what is studied and published. Breaking in is difficult for dissenters. When government agencies or corporations pay "big bucks" for science, they can determine what subjects are studied and which results become public.
A quotation that appears in many research methods texts argues that "there are lies, there are damn lies, and there are statistics." Perhaps statistics do not really lie, but the same statistics can be manipulated to defend a variety of positions.
Example: Laud Humphreys's Tearoom Trade Study
The Laud Humphreys's (Humphreys, 1975) tearoom trade study was an investigation into the sexual habits of upper-class male homosexuals. The setting was a public restroom. Approximately a hundred men were observed engaging in sexual acts. Humphreys, while playing the role of a "watch queen," followed members of the establishment to their cars. There, he secretly recorded their license plate numbers.
Humphreys later obtained names and addresses of the tearoom patrons from police registers while posing as a market researcher. A year later, in disguise, Humphreys went to the homes of the tearoom patrons to gain more insight into the lives of upper class homosexuals. To gain entry, he used a deceptive story about a health survey. Humphreys was careful to keep names in safety deposit boxes, and identifiers with subject names were burned. He significantly advanced knowledge of homosexuals who frequent "tearooms" and overturned previous false beliefs about them. There has, however, been significant controversy surrounding the study: The subjects never consented. Deception was used. Further, their names could have been used to blackmail subjects, to end marriages, or to initiate criminal prosecution. The mental anguish brought upon the tearoom patrons was severely criticized (Neuman, 1994:432).
Appelbaum, Richard P. and William J. Chambliss
1997 Sociology: A Brief Introduction. New York: Longman.
1977 "The Consumption of Dependency Theory in the United States." Latin American Research Review 12(1977):7-24.
1986 "Three scientific world views and the covering law model." In D. Fiske and R. Shweder (eds.) Metatheory in Social Sciences.
Henslin, James M.
1999 Sociology: A Down-To-Earth Approach, (4th Ed) Boston: Allyn and Bacon.
1971 "Tearoom Trade: Interpersonal Sex in Public Places." In Studies ion the Sociology of Sex. James M. Henslin, ed. New York: Appleton-Century-Crofts, 1971:351-374.
Keohane, Robert O. and Joseph S. Nye
1977 Power and Interdependence: World Politics in Transition. Boston: Little, Brown.
Mills, C. Wright
1959 The Sociological Imagination. London: Oxford University Press.
1997 Social Research Methods, 3rd Ed. Boston: Allyn and Bacon.
Reich, Robert B.
1991 The Work of Nations: Preparing Ourselves for 21st Century Capitalism. New York: Alfred A. Knopf.
Schaefer & Lamm (1992)