Empiricism Empiricism is the belief that accurate knowledge

Empiricism • Empiricism is the belief that accurate knowledge can be acquired through observation. • Important to remember: the observer does not interact with the observed. This is what is meant by 'objective observation (and measurement)'. • The scientific method is a procedure for finding truth using experimental evidence. • A theory is a hypothetical explanation of a natural phenomenon. Since however, we are going to work with hypotheses, for this class, we will consider a theory to be an intellectual framework for asking questions about the natural world. • A hypothesis is a falsifiable prediction made by a theory. • The text gives a reason for this from everyday life, but the formation of scientific hypotheses comes from the probability theories discovered by Boyle. • The first assumption is the null hypothesis, that no two (or more) phenomena are connected at all. This is counter-intuitive because neurological structures evolved to predict events, all the way back to sea snails.

Empiricism • Albert Einstein spoke about the null hypothesis this way: “No amout of experimenation can prove me right, but a single experiment can prove me wrong”. • Classical thinkers like Euclid and Ptolemy believed that our eyes work by emitting rays that travel to the objects we see. Ibn al-Haytham reasoned that if this were true, then we open our eyes it should take longer to see things far away than something nearby. With a single observation, a centuries-old theory vanished. • Scientists once believed that all combustible objects contain an element called phlogiston that is released during burning. Antoine Lavoisier demonstrated that when metals such as mercury are burned, they do not get lighter, as the phlogiston theory said they must. • The scientific method suggests that the best way to learn the truth about the world is to develop theories, derive hypotheses from them, test those hypotheses by gathering evidence, and then use the evidence to modify theories. Note that we do not set out to prove something true, but rather that it must be true in spite of contrary theories. Also no theory is the final word, it is simply the best fit to the data. • The empirical method is a set of rules and techniques for observation.

Empirical Method: Observation • The empirical method faces challenges in psychology not found in physics & biology. • Complexity: No galaxy, particle, molecule or machine is as complicated as the human brain, with 500 million interconnected neurons. • Variability: Bacteria within a species are essentially the same, whereas people are as varied as their fingerprints. • Reactivity: People often act, feel, and think one way when they are observed another when they are not. • Therefore, methods of observation had to be discovered that are unique to the discipline of psychology (and often disputed by physical scientists). • Operational Definition: concrete, measureable terms of a property. • Instrument: detect the condition under consideration (n. b. Web Article One) • Validity: the 'goodness' of fit which a concrete event defines the property • Reliability: instrument must produce the same measurement whenever it measures the same thing. • Power: an instrument's ability to detect small magnitudes of a property. • Demand Characteristics: those aspects of an observational setting that cause people to behave as they think someone else wants or expects. Fig. 2. 1

Empiricism: Measurement • Naturalistic Observation is a technique for gathering scientific information by unobtrusively observing people in their natural environment. • the biggest groups leave the smallest tips in restaurants (Freeman et. al. 1975) • hungry shoppers buy the most impulse items in a grocery store (Gilbert, Gill & Wilson, 2002) • golfers are more likely to cheat when they play several opponents at once (Erffmeyer, 1984) • most men do not usually approach the most beautiful woman at a single's bar (Glenwick, Jason & Elman, 1978) • Olympic athletes smile more when they win the bronze rather than the silver (Medvec, Madey & Gilovich, 1995) • There are ways to avoid demand characteristics: • Questionnaires can be filled out privately, with 'filler items' • Questionnaires can be anonymous, with 'cover stories' • One way to aviod demand characteristics is to measure behaviours people are unable or unlikely to control, such as pupil dilation. • Also, the orbicularis oculi muscles are not under conscious control, ie. the 'twinkle in your eye'.

Empiricism: Observer Bias • Experimenters can exhibit 'observer bias' (Rosenthal & Fode, 1963) • Some students were told that their test rat was 'maze dull'; others that theirs was 'maze bright'. The rats were actually the same breed. The first set of students reported that their rats took longer to learn the maze than the second set. And vice versa. • Another problem in the same experiment can arise with differential handling of the rats. • In order to remove observer's expectations, the double-blind procedure is used: an observation whose true purpose is hidden from both the observer and the person being observed. • Measurements are made by research assistants who do not know what is being studied or why, and who therefore do not have any expectations of the outcome. • These days computers are used for maximum objectivity.

Empiricism: Graphic Representations • Frequency distributions: graphical representations of measurement arranged by the number of times each measurement was made. (Fig. 2. 2). Did more men or more women score 50 on the fine motor skills test? • Normal distribution: mathematically defined distribution in which the frequency of measurements is highest in the middle and decreases symmetrically on both directions. • Central tendency: statements about the value of the measurements that tend to lie near the centre or midpoint of the frequency distribution. • Mode: the value of the most frequently observed measurement • Mean: average value of all the measurements • Median: the value in the exact middle. • A graphic curve where mode = mean = median is the normal curve. • A single extreme measure can skew the curve calculations, especially the mean. (Fig. 2. 5) • Range: the value of the largest measurement in a frequency distribution minus the value of smallest measurement. • A more robust measurement is the standard deviation: statistic that describes the average difference between the measurements in a frequency distribution and the mean of that distribution. (Fig. 2. 6) Men and women have the same averge I. Q. , but men have greater variability.

Empiricism: Correlation • Are two measurements (events, properties) related? Is there a relationship between A (aerobic exercise) and B (% bodyfat)? • We learn about these two by comparing the patterns of variation in a series of measurements. • Two variables are said to be correlated when variations in the value of one variable are synchronized with variations in the value of the other. • Correlations not only describe a phenomenon as it is, but predict what it will be. • So. . if I increase my aerobic exercise component by 50% (as measure in steps on a treadmill) I should decrease % bodyfat by some measureable component. • Fig. 2. 8 will be on the next exam. Which of the three correlations describes the aerobics/%bodyfat experiment? • Fig. 2. 9 gives real-world examples of positive correlations. Which of the correlations would match a theory that says: “The more beers I drink on Friday night, the happier I will get!” Pay attention to the values of r, the correlation coefficient. • Causation is something we don't deal with in psychology, unless we are doing neuroscience (that's why it is presently king). • The problem: confounding variables, or in the text's terms, “a third variable correlation” • This is why a scientific theory is never finished, once and for all.

Empiricism: Matched Samples • Third-variable problem: the fact that a causal relationship between two variables cannot be inferred from the naturally-occurring correlation between them because of the everpresent possibility of third-variable correlation • An example: Sandler's theory that polio was caused by excessive sugar consumption in children. We attempt to avoid such theorizing by using Fig. 2. 11 • Matched samples technique: participants in two groups are identical in terms of a third variable. • Matched pairs technique: each participant is indentical to one other participant in terms of a third variable • Even with these controls, we can never rule out a third variable. • This leads to experimentation, a technique for establishing a causal connection between two variables. Note how we do our best to eliminate any third variable. • Rather than observe a correlation, the goal is to create a correlation.

Empiricism: Experimentation • Manipulation: changing the value of a variable to determine its causal power • Independent variable: the variable that is manipulated by an experiment (treadmill aerobics) • Dependent variable: the variable that is measure in a study (%bodyfat) • Experimental group: the people who are exposed to a particular manipulation • Control group: the people who are not exposed to the particular manipulation • Self-selection: a problem that occurs when anything about a person determines whether they will be included in an experimental or control group. For example: what happens if I only use students who frequent the fitness centre for this experiment? • As much as possible, we would like to use random assignment (chance) for the control and experimental groups. If I only used physically fit students, would this create problems in random assignment? Why or why not? • A partial answer: using only 'fit' students from the 'fitness' centre would reduce the significance of the experimental findings. • Inferential statistics takes over here: if there is less than a 5% chance to see these particular results should random assignment fail, then the experiment is statistically significant. Note: no experimental can prove that a causal link is absolutely true; we deal with proximate truths.

Empiricism: Validity • Internal validity: everything inside the experiment is working exactly as it must for us to draw conclusions about causal relationships. We could conclude that aerobics (as we defined it) caused a drop in %bodyfat (as we defined it) in the subjects whom we studied. • External validity: an attribute of an experiment in which variables have been defined in a normal, typical, or realistic way. However, most psychology experiments are externally invalid. • Instead, psychologists try to learn about the real world by using experiments to test hypotheses derived from theories. (Mook, 1983). • A well-thought-out theory about the causal relationship between aerobics and %bodyfat about how students in the fitness centre will behave if their stated goal is to reduce %bodyfat. • Experiments are usually meant to created particular circumstances, test the hypotheses, and thereby provide evidence for or against theories that generated them.

Empiricism: Representativeness • Psychologists rarely observe a population, a complete collection of people. Why? Because it is nearly impossible to access all of them at once. Consider changing the drinking age from 19 to 18 in Saskatchewan. Could you really reach every 18 -year-old in time to ask their opinion? Would some have a birthday on the day you ran your study, so that their age status changed? Would that change their opinion? • Instead we use a sample, a partial collection of people drawn from a population. • Mostly we try to make sure that our sample is a good representative of the population we claim to be studying. Some students come from cultures that do not drink alcohol, so their inclusion would not be representative in terms of the drinking age. Or would it? • Random sampling is a technique for choosing participants that ensures that every member opf a population has an equal chance of being included in the sample. So. . . teetotallers get to stay. • By and large, subjects in psychological studies are non-random, as most are university students. And most are from the developed nations in the Northern Hemisphere. • An experimental result can be illuminating even when the sample is atypical of the population. • When the ability to generalize an experimental result is important, new experiments will use the same procedures, but on different samples. • Sometimes the similarity of the sample and the population is simply a reasonable starting assumption. The addictive pathways in the mid-brain are remarkably similar in rats and humans. (It seems rats like cocaine as much as humans do).