Experimental Psychology Chapter 6: Formulating the Hypothesis PDF

EXPERIMENTAL PSYCHOLOGY CHARACTERISTICS OF EXPERIMENTAL HYPOTHESES - Every experiment has at least one hypothesis – while CHAPTER 6:...

EXPERIMENTAL PSYCHOLOGY CHARACTERISTICS OF EXPERIMENTAL HYPOTHESES - Every experiment has at least one hypothesis – while CHAPTER 6: complicated experiments may test various ones Formulating the Hypothesis simultaneously - Experimental hypotheses are tentative explanations of an OBJECTIVES event or behavior 1. Learn the differences between non-experimental and o It is a statement that explains the effects of specified experimental hypotheses antecedent conditions – and on subsequent 2. Understand the components of a good experimental measurements of behavior hypothesis o By the time you formulate a hypothesis, you have 3. Explore where hypotheses come from thought a great deal about a behavior. You may have 4. Learn how to conduct a literature search discarded several improbable explanations of it – and are now ready to propose one explanation that seems INTRODUCTION more plausible than others - The term hypothesis has appeared a number of times in the o Before doing any experiment, it is important that you preceding chapters, as most psychological research is narrow down the possibilities and factors to be designed to test them investigated upon - Before testing them, however, it is important that we first o Once a small, finite number of possibilities have been take a look at how to arrive at them – especially in an identified – then you are ready to propose an experimental sense explanation, and state a hypothesis - Once we have arrived, we will explore how the hypothesis forms the basis of a research report – beginning with the - Every hypothesis must meet certain basic criteria – those introduction of which are far off from personal beliefs or attitudes o Just because something is true or interesting is not DEFINITION enough to make it a useful hypothesis - The hypothesis represents the end of the long process of o Hypothesis need to be synthetic, testable, falsifiable, thinking about a research idea parsimonious, and (hopefully) fruitful o The hypothesis is the thesis – or main idea – of an CHARAC- TERISTIC experiment DESCRIPTION o It is a statement about a predicted relationship between at least two variables o Some non-scientific synonyms for this are - Can either be true or false speculation, guess, or hunch - Psychology borrowed this from the field of logic - Experimental hypotheses must have some - The statement of a research hypothesis is designed to fit chance for it to be true, and some chance for it the type of research design that was selected to be false o Nonexperimental hypotheses are statements of - Experiments will be designed to test and decide predictions of how events, traits, or behaviors might between the possibilities - Non-synthetic statements must be avoided be related o Analytic statements are those that are o Experimental hypotheses are statements of always true. Analytic statements explain all predictions of cause and effect possible outcomes (e.g., I am pregnant, or I am not pregnant) or are sufficiently vague Synthetic - Some non-experimental designs do not typically include a to be true for everyone (e.g., the weight of hypothesis – particularly those that do not restrict subjects’ dieters will fluctuate) responses o Contradictory statements are those that o This is because they intend to explore and describe are always false – as their elements behaviors as they occur naturally – and it would be oppose each other (e.g., I have a brother, difficult to make guesses about behaviors or events and I am an only child) - As analytic statements are always true, and that might not occur contradictory statements are always false – o As such, nonexperimental hypotheses are there is no need for experiments to test them straightforward predictions of relationships that one - Synthetic statements are presented in the “If… expects to find between variables then” form – wherein a potential relationship between antecedents and behaviors is expressed – and the possibility is either true or false  - The means for manipulating antecedent INDUCTIVE MODEL conditions and measuring the subsequent - Induction is the process of reasoning from specific cases to behavior must exist more general principles – often used in science and math. - Many interesting hypotheses have no scientific It is the extrapolation of information observed, in order to use, because they do not meet this criterion arrive at conclusions that have not yet been observed o Ex. Do dogs dream? à If dogs display muscle twitches, vocalizations, eye o By examining individual instances, an explanatory movements – then they must dreaming framework can be devised to describe them o This hypothesis is in proper form, but can o Ex. “You must be a Libra. I can tell by your beautiful it be tested? We can manipulate clothes, your meticulous grooming, and the birthstone antecedents to encourage sleep (type of ring you are wearing” food before bedtime), but how will we - Hypotheses often come from inductive reasoning Testable know if dreams are taking place? We can o Ex. While stressing over a quiz you are about to take, ask them, but we cannot expect any useful you may have noticed athletes being called from the answers for science class – and then excused for training. You also notice o Although this is extremely interesting, it is this isn’t challenged, so you come up with the not exactly testable – as the means for observing and recording the behavior of explanation: Being an athlete allows a person interest did not exist back then privileges unavailable to non-atlhetes - Untestable hypotheses are not necessarily o This hypothesis was built through induction – wherein useless. Scientists used to speculate before several observations were made regarding behavior, what it would be like to walk on the moon – even then these instances were used to devise a general though there was no way back then to even principle to explain it travel into space - Induction is the basic tool of theory building - There is always hope that new technology will o Through induction, researchers construct theories by open new areas of discovery taking bits of empirical data – and forming general - Hypotheses must be disprovable by research explanatory schemes to accommodate those facts findings - Induction moves from Specific Observation à Pattern - As such, they need to be worded so that failures to find the predicted effect must be considered Recognition à General Conclusion evidence that the hypothesis is indeed false Falsifiable - Ex. “If you pay attention in class carefully DEDUCTIVE MODEL enough, then you will be able to pass” - Deduction is the process of reasoning from general o The qualifier “enough” within the principles to make predictions about specific instances. statement makes the hypothesis Also known as “top-down” logic à the reasoner begins with unfalsifiable – because any failures to an accepted premise, and then seeks to prove another produce the predicted effect can be statement based on previously “known” information explained away by the researcher o It is most useful when we have a well-developed - A simple hypothesis is always preferred over theory, with clearly stated basic premises one that requires many supporting assumptions Parsimonious - Ex. “If you look at an appealing photo, then your o It is then possible to deduce prediction about what pupils will dilate” is preferred over “If you look at should happen in new situations – in which the theory a photo that appeals to your experiences, then would apply your pupils will dilate if it is a warm Saturday in o Testing such predictions provides a test of the value June” or merit of the theory - Ideally, a hypothesis leads to new studies - Ex. Equity theory (Walster et al., 1978) is an excellent - Although difficult to know in advance, the example of the deductive method in psychology Fruitful fruitfulness of a hypothesis can be indicated by o Individuals will try to optimize their outcomes whether or not we can formulate new studies (outcomes = rewards minus costs) based on the experimental outcome o When individuals believe they are in an inequitable relationship, they will feel distress in direct proportion - Although there is an overwhelming number of criteria for a to the perceived degree of inequity good hypothesis, there is no fixed rule/method in doing so o The more distress they feel, the harder they will work o Despite this, there are some general approaches that to restore equity most hypotheses are formed o Equity theory has been used by researchers to predict o Understanding these can help tremendously in human behavior with great success – not only in thinking about the psychological issues you want to victimization, employment, helping, but also in love experiment upon - Deduction moves from Existing Theory à Hypothesis à Observation à Confirmation/Rejection COMBINING INDUCTION AND DEDUCTION SERENDIPITY AND THE WINDFALL HYPOTHESIS - We have looked at induction and deduction as two separate - All the approaches we have discussed thus far are approaches to formulating a hypothesis. In practice, these purposeful – the experimenter is usually looking for a new approaches are not so neatly separated hypothesis, on which to base their experiment o Their propositions were formed initially through - At times, however, a discovery can be made where none induction from specific cases was intended – this is attributed as serendipity o Later tests of the propositions were based on o Serendipity is the knack of findings things that are not predictions derived through deduction being sought o Discoveries through serendipity have been made in - Both induction and deduction are important in research – the physical sciences, and in the psychological and both are useful in formulating study hypothesis sciences as well o Through induction, we devise general principles and theories that can be used to organize, explain, and - An element of serendipity appeared in the work of Ivan predict behavior (until more satisfactory principles are Pavlov (1927) – a Russian physiologist whose main interest discovered) was the digestive glands o Through deduction, we can rigorously test the o His studies involved feeding dogs and observing implications of those theories changes in stomach secretions o Throughout his work, Pavlov became interested in BUILDING ON PRIOR RESEARCH salivation – asking the question “If I feed the dog, how - Thus far, global approaches were discussed that can be long does it take before salivation?” applied to a variety of topics. Now the process of narrowing o The questions were straightforward, but they led him down the field of possibilities will be discussed here – in onto a different path order to formulate a hypothesis o As dogs became familiar with the bread Pavlov fed them, they began to salivate – even before they were - The most useful way of finding hypothesis is by working actually fed from research that has already been done o This salivation was thus activated by sight – of the o Sometimes, nonexperimental studies can suggest food in the beginning, but also of Pavlov himself later cause-and-effect explanations – that can be on translated into experimental hypotheses o This led Pavlov to begin his studies on “psychic o Ex. Cigarette smoking and cancer à Ex post facto secretions” – which laid the foundation for classical studies revealed that smokers had higher rates of lung conditioning today cancer than nonsmokers à This suggested the o Initially, salivation was only elicited by eating the food. experimental hypothesis that “If people smoke, then After repeated pairings, however – the sight of the they will get cancer” food, and the sight of Pavlov himself elicited o Systematic manipulations of antecedent conditions salivation. This eventually was generalized over to can then be tested – wherein cause-and-effect sound as well – namely, the sound of the bell inferences can be concluded, which is not possible on the basis of nonexperimental data alone - Serendipity can be useful in hypothesis generation – but only if we are open to new possibilities - Prior experimental research is an excellent source of o Good scientists take note of all potentially relevant hypothesis observations – then analyzes and evaluates them o The more you read about a certain topic or § Are they interpretable? phenomenon of interest, the more you also discover § Do they explain something previously points that you may not have considered initially unexplained? o From time to time, published experiments might have § Does this suggest a new way of looking at the conflicting outcomes – those of which can be the basis problem? for new experimental hypotheses o Serendipity is not only a matter of luck – rather, it is a matter of knowing enough to take an opportunity as it - A thorough search of the literature available on your topic presents itself is important – both in designing a good experiment, and in writing an effective report INTUITION o Regardless of where the experimental hypothesis - Not discussed in most experimental psychology texts originates, reviewing the literature is necessary and o This is because psychology is a science, and as such essential – it should be governed by formal, logical rules o This is because one of the primary goals of report o Using intuition is not necessarily unscientific – rather, writing is to integrate your findings into extant facts the inferences drawn from it can sometimes violate scientific criteria - Intuition may be defined as knowing without reasoning o If your search is thorough, you will likely locate work o As such, it is probably closest to phenomenology – as done related to your hypothesis – or at least one that we acquire phenomenological knowledge simply by is closely related attending to our own experiences o These reports can help you develop good ideas for o We have a hunch about what might happen in a procedures to use in your own experiment – and they particular situation – so we set up an experiment to are full of tips for measuring your observations test it out o Once you have found a good article on your topic, you o Intuition can guide what we choose to study might be able to find other related publications in the article’s reference section - Still, experiments are conducted in the context of prior - Literature searches can be daunting, as there are simply research. This is done to avoid carrying out experiments too many sources available. Apart from journal articles – that are pointless – given what is already known books, expert reviews, chapter summaries, meta-analyses, - Intuition is most likely to be helpful if it comes from experts and the like can be invaluable sources of information – as good hunches are really an unconscious result of our - Generally, it is best to skip popular books written by own expertise within an area of study nonresearchers, essays from popular media, o The more we know about something, the better our blogs/wikis/personal sites, and similarly related forms of intuitive faculties can be in guiding a hypothesis nonscientific information (Halpern, 1989) WRITING THE REPORT - At all times, we must be careful to remain within the bounds - Published research reports from psychological journals will of science – whenever resorting to our intuitions form the bulk of the reading you are expected to do as o By intuition, we may have a tentative explanation for background for writing a research report behavior or events - The Introduction section of a research report consists of a o Such explanations are tentative – they cannot be selective review of relevant and recent research accepted as valid until it has been translated into a o Not everything that has ever been done must be hypothesis, and subjected to experimental testing included in your background, only those directly and o Even if we believe intuitively something is true, we immediately related to your research hypothesis must be prepared to change our thinking if the o The studies highlighted in the Introduction should experimental evidence does not confirm our belief provide empirical background for your experiment – o Observable, measurable, and verifiable data must and guide readers towards your research hypothesis always take precedence over our intuitions - Another goal of a research report is to integrate your experiment into the existing body of knowledge WHEN ALL ELSE FAILS o To show how the results of your study advance 1. Find a broad area of study that piques your interest knowledge 2. Explore deeper to further narrow down your scope of inquiry o To increase generalizability of known effects 3. Locate the latest research that exists on this topic o To contradict past findings 4. Identify gaps, problems, limitations found in those studies - In the Discussion section of your report, you will have to 5. Formulate ideas to address those identified issues refer to past work – and compare and contrast how previous 6. Refine your desired area, topic, and phenomenon of findings size up to current data and analyses investigation 7. If this doesn’t work, try observing in public places until a testable experimental hypothesis is found. This comes naturally, as we are always searching for causes behind the way people behave (causal attributions) 8. If all else fails, turn your attention to a real-world problem – and try to figure out what causes it. Once the cause is determined, the solution often suggests itself 9. Finalize your hypothesis, and set realistic goals to investigate it SEARCHING RESEARCH LITERATURE - Once you have decided on your hypothesis, the next step is to become familiar with other published studies within your area of interest o Conducting thorough literature reviews is important in conducting research – and necessary for writing reports