Plausible Rival Alternative Hypothesis = PRAH (pronounced like Ha)  This is another term for an alternative explanation (to the experimenter's explanation) for why the data came out the way it did in any particular study.

In the survey study example given the data showed tea drinkers had fewer health problems. Imagine the experimenter's hypothesis (explanation) is, "This proves tea prevents health problems."

Psychologists would have a problem with this statement. First psychologists don't use the term "prove" which is more a philosophical or math (cf. proofs) term. We would use the terms "confirm" or "not confirm." the hypothesis to be more logically correct.

But even if we change the term in the experimenter's conclusion to "confirm" from "prove there are alternative explanations for the results. A PRAH is, "This study does not confirm that tea prevents health problems, maybe tea drinkers may have a more generally healthful lifestyle and THAT is why they have fewer health problems. Tea drinking may just be an extraneous factor that goes along with healthy living like the positive correlation between drowning deaths and temperatures. They are not causally linked but tend to occur together." 

If you have done Critical Thinking programs you may have read how another explanation for the data is called an "alternative explanation". However, the advantage of using the acronym PRAH instead of alternative explanation is it is more precise and, if your class requires you to come up with examples, PRAH will guide you to better examples. You can use the initials to help develop and test the quality of your own alternative explanations. Let's work through developing a  PRAH for the tea study.

PLAUSIBLE, to say that "Tea drinkers are healthier because tea puts an energy shield around the body which repels germs."  is an alternative explanation but is NOT plausible so it is a poor quality alternative explanation.

RIVAL ALTERNATIVE, to say that tea has antioxidants which destroy disease agents in the body is NOT a rival alternative explanation to the experimenter's hypothesis. This could be the way, the mechanism by which, tea reduces disease rates in people but it is not an alternative to the experimenter's idea; it is just a more detailed description of how the experimenter's hypothesis is true. (OPTIONAL POINT >Some have complained that PRAH is a redundant phrase. The original phrase my mentor (Donald T. Campbell) used was plausible rival hypotheses (on p. 36 Experimental and Quasi-experimental Designs for Research by Donald T. Campbell and Julian C. Stanley) but that does not create a pronounceable acronym so I chose PRAH (pronounced like Ha!) as a cue to help students learn the concept and added Alternative to help students avoid a common error of coming up with mechanisms by which the experimenter's hypothesis could work instead of correct alternative hypotheses.

PRAHs are most common in survey studies, the descriptive/correlational studies designed using subject variables, like a person's gender, height, IQ, race, and so on as the independent variables.

Subject variables are subjective characteristics of the participants, personal characteristics that cannot be randomly assigned to the participants by the experimenter. The participants enter the study already having those traits so the experimenter can't assign them. They can't be interpreted causally like independent variables in true experiments because subject variables are typically associated with other variables within the participant. For example, if you did a survey study where you correlated student's grades with how close they sat to the front of the room and found the closer they sat the higher their grades you could not conclude that sitting close causes good grades. It might be that students who do well in school sit closer to the instructor as a source of past rewards and students who sit far away are avoiding the punishing instructor so it is INCORRECT to say, "Sitting close causes students to get good grades." because  it may just as well be "Students who get good grades like to sit close." This reverse causation is a common PRAH with descriptive/correlational studies where a subject variable is used as an Independent Variable.

Notice if you designed a true experiment where you assigned students seating positions randomly and you found the closer they sat the better their grades, you could accurately make the causal assertion, "Sitting close causes good grades."  By randomly assigning them to seats you have made sure the only systematic difference between the students is their sitting position so the only reason for a systematic difference in grades would be seating position.  

Subject variables are called Pseudo Independent Variables (PIVs) in our class because they look like a true experiment's independent variable but cannot be interpreted causally like a true experiment's Independent Variables (IV) can be. There are several terms used in the literature for subject variable studies like survey studies or descriptive/correlational studies. I will generally call these studies with PIVs correlational studys to help from confusing their meaning with true experiments which must have true IVs where the experimenter has assigned the participants to the experimental factor.

 These facts make for a nice mnemonic device because as a student you can more easily remember how these concepts fit together. If the study has a PIV it is a correlational study so when you see the experimenter's explanation for it you can think "HA! (an expression of skepticism and disbelief) There may be a PRAH!" Also since quasi-experiments have PIVs remember to "Mind your Ps and Qs" and you will correctly know when to look for PRAHs.

HYPOTHESIS, After studies are completed there are the
facts = what the data collected shows, and
hypotheses = the experimenter's explanation for why the data came out the way it did.
By using good experimental design, errors in the facts can be minimized.
It is very important for good critical thinking to learn the distinction between the data, the facts that all agree exist, and a hypothesis, the guess about why the data is the way it is so you know the parts of a line of reasoning that are more or less debatable.   

 Return to Internal Validity Page      Return to Home Page

Return to "tipped is the dependent variable" Page