TestableFalsifiablePreciseRationalParsimoniousThe simplest explanation is the best answerTypes of hypotheses:Causal – why did something happen?Descriptive – where, what, who1. Form hypothesis from theory, model, or observations2. Design study to test hypothesis[3. Derive predictions] ignore4. Conduct study and test predictionsGather data5. Confirm or revise hypothesis depending on observations.Confirmation: your point of view has grown strongerMust be true to your observations - no making stuff up!6. Modify theoretical conceptsLecture 13Psych 344 1st EditionExam #2 Study Guide Lectures: 11 - 27Lecture 11 (September 7)- Hypothesis must be:- Testable- Falsifiable- Precise- Rational- Parsimoniouso The simplest explanation is the best answer - Types of hypotheses:- Causal – why did something happen?- Descriptive – where, what, who- Sources of hypothesis: theories, models, own observationsLecture 12 - 6 steps of hypothesis testing:1. Form hypothesis from theory, model, or observations2. Design study to test hypothesis[3. Derive predictions] ignore4. Conduct study and test predictions Gather data5. Confirm or revise hypothesis depending on observations. Confirmation: your point of view has grown stronger Must be true to your observations - no making stuff up!6. Modify theoretical conceptsLecture 13Testing Hypotheses:•Experimenter must seek to disconfirm hypothesis.•Null hypothesis–there will be no difference between the groups.•Alternative is that there will be a difference between the groups.•We only reject or fail to reject the null hypothesis. We can never come to the conclusion that we accept the null hypothesis because there is always the next experiment which may contradict that hypothesis.Behavioral Dependent Measures:•Response rate–with humans, we don’t look at this often.•Reaction time–how long does it take you before you do the measured response.•Response duration–for how long does somebody keep up a response.Lecture 14Physiological research measures:•Galvanic skin response–palm sweat. Sympathetic nervous system.•Electroencephalogram–EEG, measures brain activity. Measure across the scull with electrodes.•Heart Rate.•Blood Tests–RBC, WBC, hormones, glucose levels, hemoglobin, iron, oxygen, cholesterol, vitamins.Types of Sampling:Simple random sampling.•All members of a population have equal chance of being selected.•Randomly draw out people. Completely anonymous.•Systematic random sampling.•Every nth member of a population is selected.•Must make sure that lists is in random order.•Stratified sampling–study of gender differences, break list into males and females and then randomly sample from them.•Cluster sampling.•Samples should be representative.•They should reflect the demographics of the population.•Subgroup of population. Closely study them as samples representative of larger populationLecture 15- Nonprobability–not random; most psych experiments done this way.•Haphazard–only okay because psychologists want to say something about the human race. So they don’t have to be very rigorous in terms of selecting the sample.•Quota–create categories and want an even number of people in each one, then randomly sample them.•Counterbalancing–depending on method, there is going to be order effects. This is how you get around them.•Complete counterbalancing.•Equal number of people in all orders of presentation.•Start half people in one room and another half in another room. Then switch them.•Can use when your study has no more than 3 conditions.•Latin squares.•Not complete, but:1. All conditions appear at each ordinal position.2. Each condition precedes and follows each condition one time ABDC; BCAD; CDBA; DACB.•Necessary when you have more than three groups.Lecture 16- Alternate forms reliability1. Administering two different forms of the same test 2. Same test but different items (ex. SAT)- Internal consistency measures (within) 1. Split-half reliability – people do roughly as well on the first half on the testas on the second half – a measurement of practice and fatigue effect. 2. Odd-even reliability – all of the answers to the odd numbered questions and all of the even numbered questions and compare if participants did as well on the odd numbered as on the even numbered questions.3. Item-total reliability – find the individual’s total score, and do a correlation on the overall test score and every section score.Lecture 17- Internal Validityo The degree to which the mathematical relationship we observe between participants’ scores actually, and only, reflects the relationship between the variables of interest—no confounding variables (you can’t explain your findings inany other way) o Cognitive psychologists are good with internal validity, not always so good with external validity • External Validity (ecological validity)o How well does the study generalize to the real world?• Content Validityo The degree to which the measurement actually reflects the variable of interesto Does your methodology measure what it’s supposed to measure? • Construct Validity – operational definitiono The extent to which the measurement reflects the hypothetical construct of interesto Can be hard to measureo Construct needs to be defined so that it can be measured (operational definition)o Did you use a proper operational definition? They never fit perfectly and always up for debate• Temporal Validityo Measurements obtained from a particular test are stable across timeo Closely related to alternative forms and test – retest reliability• Face Validityo The test is measuring what it seems to measure• Convergent Validity (i.e. Criterion Validity)o You need to look for validity in something that already exists – an already established test or measurement o Is your measure related to other measurements designed to predict the same thing? GPA, SAT, & GRE REM and Dreaming Eye movement and dreamingLecture 18• Relationships between variableso Positive linear relationships Score high one, high on anothero Negative linear relationships Score low on one, high on anothero Curvilinear function Ex. Yerkes Dodson law (relationship between arousal and performance)- Y axis is arousal, X axis is performance - High and low levels of arousal lead to poor performance, while moderate arousal leads to high performance Lecture 19o Closed-ended (i.e.