News
Is social science science?
Joseph Woelfel, professor of communication, tackled a controversial question in his public UBThisSummer lecture on Wednesday: Is social science science?
The answer, he told an audience of well over 100 attendees, is that methodologies social scientists commonly employ in studies do not qualify as scientific.
"I am a social scientist. My degrees are all in sociology. I’m proud to be in sociology. I’m happy to consider sociology queen of the sciences. But still, I say we have a problem," Woelfel told his listeners, while acknowledging that many colleagues consider his view to be "radical."
Woelfel said that despite the work of legions of thinkers, social science, as practiced today, has been unable to disprove any theory—Marxism, capitalism and Freudian psychodynamics, for instance. This is, in part, a product of the fact that the tools of social science—including inferential statistics, which help researchers decide how likely it is that data from a sample apply to a larger population—have failed to provide meaningful results, Woelfel said.
As an example of this breakdown, he cited a 15-year, $725 million study that found, through social science methods, that no significant correlation between a high-fat diet and such conditions as heart disease and breast and colon cancer. Woelfel said the scholars who drew these conclusions made a mistake: They drew conclusions based on too small a sample of data. The study included only Americans, who, compared to the rest of the world, all have a high-fat diet.
Woelfel contrasted the inadequacies of social sciences to the successes of physical sciences, noting that, through scientific methods, researchers have decisively rejected ideas, including Ptolemy’s solar system, the luminiferous ether and stars as fires in the sky.
In the physical sciences, researchers do not make assumptions about or apply the results of a study to samples they have not tested. In addition, physical scientists believe that truth is comparative—that an object cannot simply be large, but must be large or small compared to some other thing. Surveys in the social sciences that ask respondents to state their attitudes (whether the sun is very large, large, neither large or small, small or very small, for instance), provide very little in the way of meaningful information, Woelfel said.
Woelfel, in contrast to many of his colleagues, employs a method of measuring attitudes called the "Galileo system," which uses software to plot different attitudes in relation to one another, drawing on the scientific belief that measurements can only be made in comparison to some standard. The Rand Corporation, after a broad review of social science literature, concluded the Galileo system was "the closest that any social science approach came to providing the basis for an end-to-end engineering solution for planning, conducting, and assessing the impact of communications on attitudes and behaviors."
The UBThisSummer lecture series takes place at 4 p.m. on Wednesdays in 225 Natural Sciences Complex, North Campus. Remaining lecture topics include:
• June 23: "Perverted Justice: Sex Offenders and the Law," Charles Patrick Ewing, SUNY Distinguished Service Professor, Law School.
• June 30: To Be Announced.
• July 7: A lecture by Aaron Hughes, associate director and professor, Institute of Jewish Thought and Heritage.
• July 14: "From Understanding Volcanic Hazards to Preventing Their Disasters," Michael Sheridan, UB Distinguished Professor Emeritus, Department of Geology, College of Arts and Sciences.
• July 21: "Extracting Intelligence from Social Media Data," Rohini Srihari, associate professor, Department of Computer Science and Engineering, School of Engineering and Applied Sciences.
• July 28: "Toward the Bionic Human: Medical Devices and How They Are Powered," Esther Takeuchi, SUNY Distinguished Professor and Greatbatch Professor in Power Sources Research, School of Engineering and Applied Sciences.
Reader Comments
Joseph Lucke says:
This presentation, at least as reported, was a farce. Theory change within any science is far more complicated than disproof. The comment about physical scientists not making assumptions about untested samples is false. Physical scientists also claim absolute facts---the speed of light is absolute. Using similar statistical methods, studies show a correlation between fat intake and colon cancer but no correlation between fat intake and breast cancer. Which one was a mistake? Picking the Galilleo system as a pinnacle of social science research was ludicrous.
Posted by Joseph Lucke, Senior Statistician and Research Professor, 06/21/10
Jim Campbell says:
The question itself is insulting to social scientists.
Posted by Jim Campbell, Professor, 06/19/10