advertisement

Psych – Unit 1 History and Research Methods What is Psych? Psychology is a science that seeks to answer questions about ourselves – how we think, feel, are motivated, etc. It is the science of BEHAVIORS – anything an organism does (observable) and…. MENTAL PROCESSES – thoughts, beliefs, ideas, values, emotions (subjective) Psych’s Roots Modern Psychology – December 1879 – Germany Wilhelm Wundt – first psych experiment – measured sensation responses Psych = evolved from PHILOSOPHY & BIOLOGY Early Schools of Thought STRUCTURALISTS: (Wundt) wants to explore the structure of the mind. Looks inward (introspective) – very subjective VERSUS FUNCTIONALISTS: (William James) interested in how mental and behavior processes actually function and how they enable us to adapt, survive and flourish Psych’s Big Debates 1.) Nature V. Nurture Are we are the way we are due to nature (genes, chemical imbalances, genetic predispositions) or nurture (our environment, culture). Twin and adoption studies are critical….why? 2.) Stability V. Change Psychology’s Perspectives – in class 1.) Behaviorists 2.) Psychoanalysis 3.) Humanists 4.) Biopsychology 5.) Cognitive 6.) Social-Cultural 7.) Evolutionary Psych’s Subfields Basic Research: builds the knowledge base – answers content questions for the sake of knowledge Applied Research: hands on research that tackles very specific, practical problems (work place consultants) Professional Services – counseling PhD – Psychologist - doctorate in psychology – writes, researches, teaches at university Clinical Psychologist treats mentally ill MD – Psychiatrist - medical doctor specializing in psychiatry – works in hospital or psych ward or out-patient counseling Psych as a SCIENCE Psychologists use the scientific method to conduct research and produce measurable, empirical evidence/data Psych must pursue the scientific method due to the limitations of human intuition Hindsight Bias – “Monday morning QB,” I knew it all along – the tendency to believe after learning an outcome that one would have seen it coming Overconfidence – we tend to think we know more than we do (all freshmen thought they’d graduate in 4 years) 1.) Descriptive Research Methods (only describes – doesn’t explain) CASE STUDY: the study of one unit (i.e. one person) Pro – you can get detailed, in depth information Pro – great method to study rare phenomenon Con –You cannot generalize – your case study may be atypical and you cannot make assumptions about the larger population Ex: I want to write a book on the life of the American teenager and I select one 16 year old and spend a year studying that individual in detail – shadowing them, interviewing them, their friends, etc. Descriptive Research Methods 2.) SURVEY DATA – uses a questionnaire to gather information from many people quickly – often asks people to report behaviors and opinions about a given topic Pros: quick, cheap, and efficient – I can get a lot of data quickly Cons : surface level – doesn’t go in depth Cons: people lie or may misunderstand the questionnaire – or wording effect (censorship v. restrictions) Make sure your sample represents your population in order to generalize Example: I want to research the life of the American teen and have 3,000 teenagers at the local high school fill out a detailed questionnaire on their behaviors Descriptive Research Methods 3.) NATURALISTIC OBSERVATION: watching and recording the behaviors of organisms in their own, natural environment Pros – get organisms in natural setting – avoid lab effect Cons – Observer bias – two people watching the same scene may come to different conclusions (subjective) Ex: I want to research the life of an American teen, so I go undercover and enroll in the local high school. I go to class, parties, etc with the teenagers and observe their behaviors CORRELATIONAL RESEARCH Correlation studies go beyond describing behavior to predicting it Based on a known relationship we can predict phenomenon Ex: If I know a correlation exists between smoking and lung damage and I know Johnny smokes two packs a day for twenty years….I can predict that he will have significant lung damage CORRELATION DOES NOT = CAUSATION!!!!! (only a controlled experiment can show cause) How to carry out a correlation study… 1.) Identify your variables 2.) Gather data 3.) Create a scatter plot 4.) Draw a line of best fit 5.) Calculate a correlation Coefficient 6.) Analyze and make conclusions about the relationship Correlation Scatter plots Correlation Coefficients Correlation Coefficients r= + relationship: +.4 to +1.0 goes in same direction Example – height and weight - relationship: -.4 to -1.0 inverse; goes in opposite direction Example – exercise and weight No relationship: -.3 to +.3 No visible relationship Example – eye color and GPA EXPERIMENTS Experiments are the most scientifically stringent research method as you control and manipulate variables to illustrate a true cause and effect relationship IV: the variable manipulated – the treatment DV: the variable measured Operational Definitions: exactly how the IV was administered and the DV was measured Example – DV is hyperactivity. The Operational definition is a survey to be filled out by the parent rating hyperactivity on a scale of 1-10 Steps of an experiment 1.) Identify your population – the group your are interested in studying 2.) Take a large, random sample to guarantee it is representative of the population 3.) Randomly assign your sample to either the experimental condition (gets the treatment) or the control condition (gets a placebo – nothing special) 4.) Run your experiment 5.) Measure your DV 6.) Run inferential stat tests (MANOVA, t-test, ANOVA) to see if the difference between the experimental and control groups is statistically significant) – did we meet our p-value = .05? 7.) Replicate - RELIABILITY 8.) If yes, conclude that the IV caused the DV 9.) Generalize your findings back to your population Possible Confounding Variables and Critical Controls Large Random Sample Random Assignment Placebo (blind) Guarantees sample is representative and allows you to generalize Guarantees your groups are the same (=) prior to treatment Controls for the Hawthorne Effect Double Blind Controls for observer bias Confounding Variables…..BAD Confounding variables – unexpected variables that distort the results. They destroy your conclusions, so you want to anticipate possible confounding variables and control them. Controlling confounding variables helps to guarantee that your experiment is VALID – that it is truly measuring what you claim to be measuring. DESCRIPTIVE STATISTICS Descriptive stats describe/summarize a large set of data Measures of central tendency – summarized the middle of the data Mean – average Mean – middle score, 50th percentile Mode – most frequently occurring score Measures of variation – summarizes the spread of the data Range – high score minus low score Standard Deviation – average distance of each data point from the mean SD = square root of variance Distributions Normal Distributions form a bell curve. The mean, median, and mode are all the exact same point Distributions Positively skewed distributions have positive outliers – extreme scores to the right of the bulk of the data that pull the tail in the positive direction. The median is slightly affected (pulled positive) The mean is the most effected (pulled positive) Distributions Negatively skewed distributions have negative outliers – extreme scores to the left of the bulk of the data that pull the tail in the negative direction. The median is slightly affected (pulled negative) The mean is the most effected (pulled negative) Inferential Statistics Inferential Stats are more sophisticated statistical tests (t- tests, ANOVAs and MANOVAs) used at the end of an experiment to infer significance P-value = .05 If we meet the p-value we are 95% confident that the difference between the experimental and control groups is caused by the IV and is not due to chance (i.e. only 5% chance of error)