Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Is Psychology and social science really a science?

  1. Mar 30, 2010 #1
    Maybe it's just me, but it seems that more often than not people in the fields of psychology and social sciences will abuse the scientific method in order to justify their own inane BS. Am I right or wrong? Please explain, thanks.
  2. jcsd
  3. Mar 31, 2010 #2

    Math Is Hard

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Could you share some examples?
  4. Mar 31, 2010 #3


    User Avatar
    Gold Member

    I think its fairly universal to justify things using a sort of scientific method. I practically identify the word "justify" with the idea of "scientific method" even.
  5. Mar 31, 2010 #4


    User Avatar
    Homework Helper

    I wonder if this applies today.
    Last edited by a moderator: Sep 25, 2014
  6. Mar 31, 2010 #5

    Char. Limit

    User Avatar
    Gold Member

    Psychology is somewhat of a science... yet sociology is just a backdrop for propaganda.

    Sorry to burst your bubble.
  7. Mar 31, 2010 #6
    I reckon that about 5% of the papers in psychology meet the scientific method. For the most part it's just 'any-thing goes' reasoning. It's also highly influenced by current cultural opinions. Read this:

    Psychiatrist Fredric Wertham asserted in his 1954 book Seduction of the Innocent that "Batman stories are psychologically homosexual". He claimed, "The Batman type of story may stimulate children to homosexual fantasies, of the nature of which they may be unconscious" and "Only someone ignorant of the fundamentals of psychiatry and of the psychopathology of sex can fail to realize a subtle atmosphere of homoeroticism which pervades the adventures of the mature 'Batman' and his young friend Robin."

    (Wertham, Fredric. Seduction of the Innocent. Rinehart and Company, Inc., 1954. pg. 189–90)

    Now, then you can say 'Yeah, but that was then', well, I'm pretty sure bullshιt still happens. A lot of 'child psychologists' of today will go out and claim that for instance swearing a lot at home or certain media is 'bad for the development of a child', however they failed to define 'bad for the development of a child' in hard terms that can be testable against research, and I've really never seen any scientific study with controlled conditions that backs up this view that if children hear words like 'fυck' or see people exercise the act linked to that word on television they some-how become mentally compromised of some sorts, for one, such a study would be unethical, what if you're right? then you've just compromised half your group of children.

    Psychology for the most part is just spewing out theories that sound 'intuitive' to you and your colleagues without backing them up with experiments. Also, there are different 'schools' of psychology which often claim the inverse from the same evidence. Psychology is for the most part a pseudoscience which is as vague as astrology, there is no more evidence for types like meyers-briggs then there is for enneagrammes. Stuff that supposedly men like womwn that half large breasts and an hourglass figure is because that indicates fertility just gets accepted in journals there. Well 1: that's nurture, not nature, it's just a western cultural thing that's already well over it's peak, a lot of men already like an A cup. And 2: There hasn't been any indication whatsoever that women with larger breasts and thighs randomly are more 'fertile', as in, they can have more children, and 3: Then you still have to show that it's because of that.

    Psychology is observing things which may not even be correctly observed, and then trying to concoct a nice intuitive explanation for it without the rigours of experimental verification; it's also for a large part just trying to justify popular dogma and myth with a 'scientific allure' to it.
  8. Apr 10, 2010 #7
    Psychology doesn't really seem that scientific as of now (compared to more experimentally verifiable studies), but don't demean the whole thing. It's still young and it's gotta start somewhere right?

    It's not like Newton said "When I throw this apple up into the air, the space-time curvature caused by both the apple and the Earth causes an apparent attraction between the two".
  9. Apr 28, 2010 #8
    I'm more into physics and many of the natural sciences, but I think psychology sounds interesting, so maybe I'll take a shot at this. The reason I care about this topic is everything around us is affected by how we think/act, so finding more universal patterns in the way Isaac Newton looked for patterns would be very beneficial. All of Science, which is useful for society, is also affected by our problem solving abilities, so that another reason I think studying psychology and making it more of a hard science would be helpful. Psychology is definitely more soft science compared to biology, chemistry and physics, However, from what I can tell academic psychology definitely uses the scientific method, falsification, peer-review method. To put it into context, I'm not talking about media psychology, but rather the academic peer-review territory.

    Some food for thought here, if you have two groups which are the exact same and the only difference between the control and the experimental groups is an independent variable, then it would seem there's a cause-effect between that and the dependent variable. Furthering this line of thinking, although you can't directly observe thought process, you can observable behavior, which if a type of experiment can be repeated/verified over and over again then there may be something there.

    Along with that, keep in mind the natural sciences study many things we've never directly seen, but allow into the Realm of Science because they're falsifiable and you can go with the explanation that fits the evidence the best. In biology they say endorphins are one of the body's natural pain killers. They'll give an experimental group a shot of endorphins and a control group a placebo, then give them a scale to rate their pain and when there's a statistically significant difference most will say that's scientific. Then they'll observe people naturally have less endorphins in them when they're experiencing more pain. "PAIN" in this situation can't be observed directly, but rather only indirectly. No matter how much they use brain scans, that still doesn't mean proof that pain exists in people other than you as an individual because all biologists/neurologists can observe is there's a correlation between brain scans and whether they act/claim they have pain. So I'd think psychology could be compared along those same lines, even if more of a soft science, but if they make falsifiable hypotheses in peer-review journals than that would be considered the scientific method. Likewise, you know in physics we have electrons and no one has any real life pictures of them, but that doesn't mean it's not reasonable to believe in them because they've been tested so much from so many angles.

    I'd say psychology is a science with a real life scientific method, even if not quite as hard a science because it's not what Kuhn would call developed into the paradigmatic stage that chemistry and physics are.
    Last edited: Apr 29, 2010
  10. Apr 28, 2010 #9
    I think you need to differentiate between the many different types of psychology. Neuropsychology is practically biology with some ee for good measure, and therefore has some pretty hard core experiments, as does some of the cognitive and behavioral stuff. Epidemiology, lots of social psych, and a good chunk of sociology, is pretty much applied statistics, which means that the papers are as valid as any other fields that rely heavy on statistics (which is most of them.) Then there's psychotherapy, which many of the neuro/cog/behavioral/stat/etc. psychologists hate precisely 'cause it is a field that's rife with unverifiable/untestable claims. It's slowly transitioning into a fringe element of psych, rather than a major school. Even a lot of clinical psych programs are moving away from Frued, Jung, and anything else that seems very unscientific.

    EEG experiments run with about a 100 or more trials are pretty standard in cog-neuro experiments, though the presence or control and experimental groups depends on whether the experiment needs it.
  11. Apr 29, 2010 #10
    To put into context what I was trying to say about that, basically I talked to a biologist once who said psychology isn't a science because in biology you can directly see things like animals, leafs, cells, etc, but you can't directly see thought processes, even if they do use experimental-control studies. So that's why I made the mention that you can observe observable behavior in experimental-control studies, and that you can make many types of cognition "falsifiable". In areas of the natural sciences there are things we can't study directly, but are rigorously falsifiable through indirect observation and we can say "this is the best explanation for all of the evidence". Although the various areas of psychology are different, I'd say it depends on how they use their scientific method.
  12. Apr 29, 2010 #11
    Something to think about as far as figuring out if peer-review psychology uses the scientific method:

    Have you ever had a time when something bad happened and someone else said, “You should have known better. It was obvious what would happen. I don’t feel bad for you," then you said there was absolutely no way you could have known it would turn out that way, and the other person kept on insisting? One side may argue people often use hindsight bias to judge others unfairly after the fact. However, others may argue that people are really good at telling if someone should have known better after the fact. As far as why we should care about what the Scientific Method may find out about this, there can be follow-up studies on how to deal with these situations and implications which may be found useful for society.

    Just like in the Medical Field where they have an experimental and control group to control for third variables, I found it interesting the same thing was done to test hindsight bias in victim blaming. There were two groups of people, both whom received the same exact account of a man and woman who met at college and then went on a date, the only difference was the experimental group had an ending sentence saying the woman was raped.

    Experimental group’s last sentence as independent variable:

    “The next thing I knew, he raped me.”

    Control group’s last sentence as independent variable:

    “The next thing I knew, he took me home.”

    Then the participants were given a questionnaire asking them to pretend they didn't know the ending and to judge from the details of the account how much she was setting herself up for rape. I thought the results were interesting, plus I've seen it replicated again and again in different ways:




    If there are two groups the same, with an independent variable different, then that means there's a cause-effect relationship. Although you can't directly observe or prove 100% many things in Science, you can make it falsifiable/testable and go with the best explanation. At the very least you can directly observe that everything was the same except for the very last sentence independent variable, and the dependent variable of verbal behavior. Outside of the lab many have directly observed how people after the fact say, "She should have known better." The point of this study was to use experimental-control to see if there was a cause-effect by controlling for third variables. Although it in no way proves someone is innocent, there does appear to be cause-effect of hearing the ending and verbal behavior of how likely something was to happen (and verbally putting together the details to back oneself up in studies if they ask "why you say it that way"). Just like the natural sciences, there can be follow-up studies and falsification.

    Peer-review source I used:

    Janoff-Bulman, R., Timko, C., Carli, L. (1985). Cognitive Biases in Blaming the Victim. Journal of experimental social psychology, 21(2), 161-177.
  13. Apr 30, 2010 #12
    Man, you're hanging out with the wrong psychologists if you think it's surprising that researchers would use science to test claims. The kind of study you mention isn't a novelty, it's standard and has been for ages.

    Neuro-cog people do the same thing using EEGs, fMRIs, and a host of behavioral tests. It's really not all that different.
  14. May 1, 2010 #13
    I think you're understanding me out of context. That last post I made with the graphs was to the many posts on this thread which kept on saying psychology isn't science (one poster said only 5% use the scientific method). I've taken academic psychology courses which use peer-review and have looked at quite a few peer-review myself, so I just simply gave those on this thread an example to think about where an experimental-control study was used. I couldn't just say, "This is the way it is", but rather decided to be more indirect, "This is some food for thought with a reference for the graph I made." I like to treat people like independent minded intellectuals.

    If you look at peer-review journal articles for physics, chemistry, biology, and psychology, it's obvious that physics/chemistry are more hard science versus soft when compared to psychology. However, it's obvious that academic peer-review psychology definitely uses the scientific method and so can be considered "scientific" even if not hard exact science like physics/chemistry. It's also obvious there are many who claim to be psychology but don't use the scientific method and are media psychology, so that's why I made the distinction earlier that it depends on the method they use. Even for academic peer-review psychology, there is a hierarchy of sciences where there are more hard vs. soft psychology in how scientific they are, just like there is with the natural versus social sciences. Psychology would also be what Kuhn may call not as paradigmatic as physics/chemistry; many would say there's a continuum of hard vs. soft facts using the scientific method, rather than black-white.

    I just wanted to put what I was saying into context, since I'm not the best at doing that sometimes. Sorry for any confusion.

    When I mentioned above "Furthering this line of thinking, although you can't directly observe thought process, you can observable behavior, which if a type of experiment can be repeated/verified over and over again then there may be something there," the reason is biologists are sure to say that experiment above with the graphs isn't like biology where you can directly observe mitochondria, leaves, etc. The point I was trying to make is although that's true, many areas of the natural sciences study things not directly observed and make abstract things at the very least falsifiable and then go with the explanation that fits the evidence the best.
  15. Jun 4, 2010 #14
    The mind is like a computer. You have software (ideas) and hardware (brain). There is a fundamental difference between the two. Trying to understand behaviour in terms of ideas or motivations is a completely different story in comparison with trying to understand behaviour in terms of the things which are possible or not possible in terms of the make up of the brain. In neither of these two different realms a formal theory has come into existence. As far as I know the only exception might be inhibition theory developed to understand the consecutive reaction times in prolonged work task, go to:


    Inhibition theory would be an example of a formal theory of the second type of explanation. Physics as a formal science came to existence with the development of Newton's theory of gravitation (classical mechanics). The birth of a formal theory is the starting point for a science. Since in neither of these realms a formal theory has come into existence one may conclude that Psychology as a science does not yet exist. If such a theory would exist then the theory would have a name as well as the name of it's originator. Everybody knows Newton's gravitation theory or Maxwell's electromagnetic theory or Einstein's relativity theory. Do we have the name of such theory in Psychology? The answer is no.

    By the way, explaining behavior by ideas is a matter of means and ends analysis, explaining behavior by the conditions of the brain is a matter of cause and effects analysis.
  16. Jun 5, 2010 #15
    A lot of sociology is propaganda-type knowledge that spins concepts in a way to promote certain agendas. For example, sociology textbooks will note that spending on education is less than other spending and associate that with some other social problem, but the assumption that is taken for granted is that more spending on education automatically improves the quality of education and social life in general. It would be more scientific to investigate how education funding affects education processes, but that doesn't allow for neat statistical correlations.

    The sad thing about sociology is that in an effort to gain scientific credibility, many sociologists have embraced complex statistical methods that make their research appear to be rigorous and objective when in fact what these models do is distract from critically reasoning about the theoretical assumptions in the model and background theory. Naive critics will claim sociology is less scientific if it remains purely theory-based and qualitative, and this puts pressure on sociology to become less rigorously scientific by addressing relatively theoretically weak questions with complex statistical procedures - because this satisfies political and social institutions that seek to promote their agendas with sociological research. Scientism in sociology IS the reason it is propagandistic a lot of the time.
  17. Jun 6, 2010 #16
    I cannot agree more.
  18. Jun 16, 2010 #17
    Lots of social science is politically motivated just as some hard science is. When statics agree across multiple sources with reasonable sample sizes amounts intuitions and journals with no obvious agenda's then we can feel better about their conclusions. We can also feel better about the conclusions when the article has been in a journal for a while and has had time to stand up to criticism.

    One important point in the scientific method is reproducibility. If we feel a result is biased we can examine there method and try to reproduce their results. We can also suggest improvements to the method which we feel will better remove bias.

    Just because there is the potential to abuse and lie with statistics does not mean we shouldn't look for hard evidence and numbers to try to support and refute our conclusions.
  19. Jun 17, 2010 #18
    I think it would be enterprising to set up an experiment in psychology/sociology that would measure (be an appropriate definition) the degree of subjective bias that enters into conclusions of psychology/sociology. To this end, I would propose that the experiment should use Peer Expectations as the biasing agent. That is, the experiment would examine how published papers, say, are bias is such a way as to be judged worthy of positive references by others who also publish upon the same subject.

    Is anyone getting the idea?

    Andre has called this groupthink, and that's appropriate. (If you disagree on my interpretation of your meaning, Andre, let me know.) Those who fail to fit within the group will find themselves out of the group and out of a job in their profession of choice. And even before this, academic chances are diminished.

    Anyone foolish enough to take up such an experiment, attacking one's support group, and truthfully publish the results, will find him/herself in similar straits. Been there; done that, many times, in similar circumstances.
    Last edited: Jun 17, 2010
  20. Jun 17, 2010 #19
    Max Weber is an early pioneer of recognizing and controlling for political and personal interests in research. Some people take his advocacy for value-freedom very superficially and think it it refers to using statistical methods that are supposed to control for researcher bias. In reality, statistics designed to appear more objective are often actually more prone to bias in that they amplify the interests behind the research questions in the first place. For example, a statistical study comparing average IQ's according to gender or ethnic categories is already biased in the direction of comparing people along those lines as opposed to, say, educational attainment, job sector, or what color car they drive. Likewise, it would contain the bias that IQ accurately measures intelligence and that average IQ of a category reflects something inherent among individuals classified as such.

    I don't believe that Weber or contemporary proponents of subjectivity-research would say that bias can be eliminated from social scientific research. Instead, they would promote openness to recognizing, discovering, and disclosing biases and interests present in research. In this way, readers can themselves control for the bias by considering what interest there is in conducting certain research or publishing certain claims/findings. The simplistic response, common in public discourse imo, is to totally invalidate any research that appears to be the product of interests or contain bias - but in reality this just gives more power to the researchers who do the best job of concealing their interests and bias and giving their research the most convincing aura of objectivity.

    What it comes down to, though, is that the best way to discover bias/interests in research is to analyze the (potential) effects is has or will have in discourse. Even when researchers themselves are unaware of the implications of their research or are unintentionally biased in some way, their publications can have discursive effects as if they were. This doesn't mean that the researchers are responsible for these effects, at least not necessarily or soley. Discourse involves complex interactive meaning-making and spin can take place just as much at the level of the reader as the author. Then you have to consider that in academic discourse, the reader is often using literature to ground subsequent research, so there is a lot of complexity to be mindful of.
  21. Jun 18, 2010 #20
    Nowhere do you mention cooking data, which I find very common--today even, and most rewarding. It's a lot easier to start with a goose than decorate a chicken to look like a goose with words and statistics.

    Then there are cases which use no data at all, and are purely gleeful imagination.

    In psychology we have hypnotic regression.
    In sociology, my favorite is the Social Text/Alan Sokal exposure.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook