Is Psychology a Science?

What is science?

According to BF Skinner (1980), ‘there is no place in a scientific study of behaviour for a mind or self’.  But, pick up any A-level psychology text and it will be described as the study of mind and behaviour; often in the title.   

Burrhus Frederic Skinner

Others go further and believe that even an observable characteristic such as behaviour cannot be studied objectively and certainly not when it’s human behaviour. 

To what extent therefore can psychology claim to be a science?

Modern definitions of science

Recent attempts to define what makes science scientific have generally included the following characteristics:

1. Controlled observations

Generally in scientific research something (the IV) is manipulated and we observe the effect this has on something else (the DV).  A physicist might manipulate the weight of a pendulum and measure its period, whilst obviously keeping length of string and height of release constant. 

In psychology this characteristic is best exemplified by the laboratory experiment where as many variable as possible are kept constant to see if the IV is causing the DV.  

2. Objectivity

Physics and chemistry are objective and hopefully mostly free of personal opinions but is psychology?  Popper demonstrated this to an audience of students.  He said ‘observe.’  After a pause the reply was ‘observe what?’  Popper had made his point.  When we observe we look for certain things.  In research we set out looking for certain behaviours or characteristics.  We have a predetermined idea of what we’re looking for and as we all know if we set out with expectations we’re quite likely to meet them.  Essentially this is the argument of the social constructivists.  Our existing knowledge determines our expectations and our viewpoint.  This is particularly noticeable in psychology with researchers belonging to one approach or another, e.g. cognitive or behaviourist.

3. Testing theoretical predictions

Having created models or theories we are able to make predictions based upon these.  We can then test these predictions with research.  Work on spatial memory in meadow moles produced the spatial adaptation model of animal memory.  Subsequent research on lizards showed this to be wrong resulting in the pliancy model.

4. Falsifiability

A concept introduced by Popper in 1969.  Having a theory that can be objectively tested and ultimately proven wrong is what distinguishes science from religion and pseudoscience such as psychoanalysis.  Psychological research tests an alternative or experimental hypothesis, however, we are not seeking to prove this, rather we seek to disprove our null hypothesis.  As Popper put it;

‘No amount of observations of white swans can allow the inference that all swans are white, but the observation of a single black swan is sufficient to refute that conclusion.’

In psychology many theories have been tested over the years and been shown to be wrong.  Weiss’ replication of Brady’s ‘executive monkeys’ experiment highlighted a crucial error in the research.

Similarly Schacter and Singer’s controversial research of 1962 was tested and questioned on methodological grounds seventeen years later!

However, there are concepts in psychology that do elude testing and falsifiability.  Freud’s hypothetical constructs, the Id, Ego and Superego, along with Eros and Thanatos and the psychosexual stages and the Oedipus complex can never be tested objectively.  They will forever remain non-falsifiable.  Similarly Maslow’s hierarchy of needs.

5. Replicability

Already mentioned on numerous occasions above; others must be able to test your findings.  Generally in psychology, laboratory experiments can be repeated, provided sufficient detail is included in the published article.  Much of the behaviourist approach has been tested many times and the schedules of reinforcement for example are seen as close to psychological fact as it’s possible to get.  Piaget’s work has been tested to death too!

Replication in social psychology however is more hit and miss.  Generally, research such as that of Asch and Milgram that was set in tightly controlled environments has been repeated.  Research reliant on real life observations is not always so easy to recreate. 

In order to replicate research all details need to be included in the write up, including details about participants, procedures, design decisions and of course the raw results.  Sir Cyril Burt’s research (below) is a lesson on what can happen when research is not published in full. 

On a note of historical interest it is worth pointing out that Burt’s work formed the basis of the eleven plus system in this country, which I missed by one year.  Testing at age 11 determined whether you went to grammar school and got a good standard of education or were relegated to a secondary modern school and got a poorer standard.  According to the argument, since intelligence was largely genetic then being thick at 11 meant that you’d always be thick so there was no point in wasting a good education on bad genes!

Working with 21 pairs of MZ twins reared apart in 1955 Burt obtained a correlation in IQ of 0.771.   Eleven years later, and now working with 53 pairs of MZ twins Burt’s figure for correlation was remarkably still 0.771.

Leon Kamin was the first to question the findings and subsequently the co authors who Burt claimed to work with could not be found.  His work could not be verified since his wife burnt it on his death!  This led close friend of Burt, Leslie Hearnshaw, who had read the eulogy at Burt’s funeral and had been chosen by Burt’s sister to write his biography, to conclude that much of Burt’s research had in fact been fraudulent.

6. Paradigm

According to Kuhn, a paradigm is the most important aspect of a true science.  Essentially a paradigm is a framework or central concept around which the science fits.  For example the Laws or relativity for physics.

As a science emerges and develops it progresses through three distinct stages:

Pre-science

No clear paradigm has yet emerged.  There may be a central thread holding the subject together but it is a broad church, perhaps embracing many different theoretical perspectives and as yet unclear as to which one to follow.

Normal science

A widely accepted view has emerged that seems best able to explain current observations in the field.  For example Newton’s Laws of Motion (prior to Relativity).

Revolutionary science

A sort of transitional phase to a new paradigm.  As evidence accumulates and new observations emerge the old paradigm may look increasingly shaky.  Gradually thinking in the area starts to gravitate (speaking of Newton) to a newer paradigm.  However, this tends to be a slow process as many of the proponents of the existing paradigm refuse to give way.  Eventually however, there is a paradigm-shift as the new paradigm acquires concensus.  An example would be the movement from the geocentric to heliocentric view of our solar system.

Applying paradigms to psychology

I think you can see where this one’s leading.  Re-read the three stages above.  Where does psychology appear to be?  I think the clue is in the phrase ‘different theoretical perspectives’ which seems to sum psychology up quite nicely. 

According to Kuhn (1962) psychology is in the pre-science stage.

1. There is no one central approach, rather a collection of different theoretical ideas centred around psychodynamic, behaviourist, cognitive and humanistic thinking.
2. Because psychology covers such a huge area it tends to tread on the toes of those around it.  It overlaps with biology, sociology, neuroscience, philosophy and so on, which have little in common with each other. 

Of all the theoretical approaches it is behaviourism that comes closest to having a paradigm since it is simply a study of behaviour with the idea that everything is learned. 

However, you psychologists feeling put down by not being seen as scientists (by Kuhn at any rate) take heart.  Many of the physical sciences struggle to produce a paradigm.  Chemistry for example sub-divides itself into organic, inorganic and physical each with different assumptions and approaches.  

Peer review (refereeing)

This is seen as essential in scientific research and again relies on thorough and accurate reporting of scientific research.

The process doesn’t usually start until a report has been submitted for publication in a scientific journal.  Prior to publication the editor will ask external scrutineers to look through the journals that have been submitted and essentially pick out the best.

Research is only published if:

1. It makes an important contribution to scientific knowledge
2. It has sound methodology
3. It is ethically sound

However, there are important problems with the process that may lead to bias.  Imagine you’re carrying out the process.  What will impress you most, an original study or a replication?  A significant finding or one where firm conclusions cannot be drawn?

Non-significant findings and repeated studies produce the ‘file-drawer problem.’  They never see the light of day and result in a skewing of scientific publications.

Generally however, the process is seen to ‘add merit’ to the scientific approach.