Exam 1 Research Design PDF

Chapter 1: Introduction to Psychological Research From the Greek words psyche, meaning “mind”, and logosmeaning “the study of” The study of behavior and mental processes But how do we study it? Use our intuition? Keep a diary about our feelings? These are subjective and pro...

Chapter 1: Introduction to Psychological Research From the Greek words psyche, meaning “mind”, and logosmeaning “the study of” The study of behavior and mental processes But how do we study it? Use our intuition? Keep a diary about our feelings? These are subjective and prone to bias. We should use a scientific approach. How do we study psychology scientifically? To answer this, we first need to answer: What is science? Science is a system for acquiring knowledge, through observation and experimentation Science is not simply a collection of facts, but a process for interrogating the world. Facts are the product of science. Science as a Process of Discovery Two components EMPIRICISM SKEPTICISM EMPRICISM Claims based on evidence (i.e., data) Data derived from research using observation and/or experimentation Claims based upon personal anecdotes, experiences, and/or intuition do NOT equal data Skepticism Skepticism also is valuable when evaluating the source of a claim Pseudoscience Has the appearance of science, but is not scientific because it lacks a reliance on empiricism and skepticism How well does this describe you? Davies (1997): participants took a personality test, and then no matter how they responded, received the above paragraph as the results Finding: Putting pseudoscience to the test…  In sum... Pseudoscience in Marketing Phiten Scientific literacy is incredibly important. And not just for scientists! So… Are psychologists scientists? Depends on the type of psychologist Research psychologists Clinicians that rely upon empirically supported treatments… not as common as you might expect Psychologists are scientists if they rely upon empiricism and skepticism An updated definition of psychology: The scientific study of behavior and mental processes This is what separates it from other approaches such as literature, philosophy and religion Psychology as a science is relatively new (late 19 century) th This class is all about learning the scientific methods underlying the field of psychology “But Psychology is just common sense!” For example: But wait a second… – You will be better prepared to contribute to the field of psychology ▪ Either as a researcher, or as a practitioner – You will become a better evaluator of information​ ▪ This will help you make more informed decisions – Other psychology classes focus on what we know about psychology This class focuses on how we know these things Moving beyond being a psychology “appreciator” – Become an informed evaluator of psychology – And even a creator of it  – This class is the equivalent of film school for movie-lovers: ▪ learn how the thing you love is made, and how to make it yourself Chapter 2: Introduction to the Scientific Method Where do research questions in psychology come from? Every day we all ask psychology questions! ▪ Examples from my life What are research methods? Why are they important? – You also often employ research methods w/o realizing it – Example: my fuzzy-headedness last week Think about any of the famous psychological studies you’ve read about. – How did research methods contribute? Why use the Scientific Method? To understanding the world through empirical evidence rather than intuition Scientists are guided by two general precepts*: 1) The world is intelligible – With work, we can reveal knowledge and understand its patterns – “It just is”, “it’s magic”, etc. are not explanations. Understanding is the goal. 2) This process is hard. Our minds are prone to illusions and fallacies – Sources of belief like ideology, conventional wisdom and subjective experience are prone to error – We must correct for our biases with skepticism and empiricism *Steven Pinker, “Science is Not Your Enemy” (8/6/2013) http://blogs.tnr.com/article/114127/science-not-enemy-humanities Conclusions based on evidence Evidence is examined by studying how variables are related Variables are elements investigated in studies that are manipulated, compared, and/or controlled Examples: gender, amount of TV watching, level of depression Independent Variable (IV): the predictor variable. Often systematically altered or manipulated by the experimenter  Dependent Variable (DV): the outcome variable. It’s what is being affected by the IV. Examples: 1. 2. 3. 4. What are the goals of applying the scientific method to psychology? 1. Description 2. Prediction 3. Explanation 4. Application Goal 1: Description Questions description addresses: – What is the phenomenon? – What are its primary characteristics? – How do we know when it is present? – For whom does it occur? – Under what circumstances? Example: what is hunger? A more difficult example: what is love? Goal 2: Prediction What is associated with the phenomenon? Are two things related? – Example: One variable that can predict another means that they are correlated – Examples: – Correlation doesn’t tell you why the two variables are related Correlation Correlational Observations: Birth Control by the Toaster Method Correlation mistaken for Causation The most common error when interpreting scientific findings Goal 3: Explanation  Why does something occur? – What causes it? not just what’s associated with it Causation: when a second event is the consequence of the first (i.e, one variable caused an effect in the other). – Experiments determine causality by eliminating confounds (i.e., other variables that may be causing the effect) How are correlation and causation different? – Prediction = correlation – Explanation = causation Example of a Simple Experiment Does reading before class increase class participation? Goal 4: Application How can this information be used to affect the phenomenon? – Applied research vs Basic research Can even basic research be applied to improve lives? How? The Research Process Step 1: Ask a question A research question should be – Relevant and Important – Contribute to knowledge ▪ not waste your and the participants’ time and resources Example of a less relevant (but totally testable) research question: – Do kids tie their shoes faster if they drink OJ immediately prior? Much more relevant, with practical implications: – Are preschoolers who use math flashcards better at math once they are in elementary school? Step 2: Study the Literature What is known about this topic already? – Review existing research literature and theory critically – PsycINFO and other internet search tools http://www.utdallas.edu/library/ ▪ Click “Find Articles & Databases” on the right ▪ Click “P”, then “PsycINFO” – Use previous research to generate new hypotheses  ▪ Should generate new knowledge (although replication is important too) Step 3: Develop Hypotheses What is a hypothesis? – A testable prediction about a phenomenon – Has at least two variables and specifies a relationship between them – May provide a causal explanation, or just a correlation Produce a hypothesis – What is your IV? – What is your DV? Step 3: Develop Hypotheses, ctd After reviewing the literature, ask: What do we know, what do we know well, what don’t we know? – What are the problems with previous studies? – What has not been tested? With whom? – Do current theories sufficiently explain the phenomenon? A Hypothesis is: Testable To be testable, constructs must have an adequate operational definition Construct: a psychological concept or idea ▪ What are some examples of constructs? Operational definition: defining constructs in terms that can be objectively measured A testable hypothesis makes a prediction – Not a hypothesis: Self-esteem in teenage girls – Better, but still not a great hypothesis: – A hypothesis: A Hypothesis is: Adequately Defined Included constructs must have operational definitions. Example: Better: Example: Intelligent people are happier Better:  A Hypothesis: Cannot be Circular Example: People who are highly aggressive will be more likely to hit another person Better: – Example: Children with ADHD will pay less attention during a task. Better: A Hypothesis is: Falsifiable To be scientific, an assertion must be open to refutation (i.e., capable of being proven false). Examples of a falsifiable assertion: Women are better than men at recognizing emotion; early hearing loss impairs language development; students who read the textbook perform better on the exams than those who don’t, etc. Examples of assertions that are not falsifiable: Morning sickness results from unconscious hatred of the father; stealing is wrong; rock music is better than jazz; guardian angels exist; all are not scientific claims A Hypothesis must Involve: Ideas Recognizable to Science Terms need to be operationally defined and falsifiable to be scientific. Example: People commit crimes because they are morally inferior and open themselves up to sin Better: Example: People who God blesses are happier Better: The difference between the two is that the second is testable and open to being proven false Reminder: A good hypothesis offers a relationship between variables that is supported by previous research and/or theory A hypothesis may originate from feelings, personal experience, or intuition… but these are often unreliable, or have already been tested. – A hypothesis must be grounded in previous scientific evidence! Step 4: Design a Method  Once you have a hypothesis with operationally defined constructs, choose your measures: – Reliable, valid measures with previous support ▪ Reliable = how consistent a measure is ▪ Validity = does it measure what it’s supposed to? The “truthfulness” of a measure ▪ Both reliability and validity have various sub-categories that need to be fulfilled. We will discuss these later ▪ External Validity = are the results applicable to the real world? – Are these measures reliable? Valid? ▪ Test someone’s intelligence by measuring his weight ▪ Measuring weight with a scale that’s randomly inaccurate ▪ So something can be reliable but not valid. However, it’s impossible to be valid w/o reliability. ▪ Step 4: Design a method, ctd Qualitative (ex: case study; interviews) vs. Quantitative (research translated into and analyzed as numerical data) Correlational vs. Experimental vs. Quasi-Experimental Specify a Sample Who are you going to test? – You can’t test everyone, so you’ll need to use a sample. The sample should representative. – If possible, use random-assignment in your sample How many participants do you need? – Power to detect whether the IV affected the DV Step 5: Collect Data Recruitment Ethics - We will discuss next lecture Train staff to administer the tasks Get subjects to participate, fully and honestly Step 6: Enter and Analyze Data Descriptive statistics: summarizing (e.g., describing) your data – How did participants do on your tasks?  ▪ Example: men recognized 78% of emotional expressions correctly while women recognized 82% – Measures of central tendency (e.g., the “average”) and variability ▪ We will discuss these more in depth later Step 7: Interpret the Data Was the hypothesis supported? Use inferential statistics to determine. Inferential statistics: testing for meaningful (i.e., significant) group differences – Did manipulating the IV have a meaningful effect on the DV? ▪ Example: is the 82% correct for women meaningfully different than the 78% for men, or is it likely that the difference just happen by chance? ▪ Or, did a new pill (IV: one group got it, the other got a placebo) help to reduce a cancerous tumor (DV)? Step 8: Disseminate the Results Present and/or publish the data – Peer review – Quality of outlet More on “communicating research” later Remember: A single study is insufficient to answer any research question – hypotheses and theories are supported by replication Highly replicated findings converge to support a scientific theory, which is the pinnacle of knowledge – The number one sign of scientific illiteracy: misuse of the term “theory”. – The “Theory of Evolution” does NOT mean that this is a guess about what might be true. It means all the known data lead to this conclusion. A “theory” in science is the most rigorous, well-confirmed and reliable form of knowledge. It is not just a “guess”. Ethics in Research MILGRAM EXPERIMENT Many aspects of the Milgram Experiment have since been deemed unethical, and likely would not be allowed in current studies But were they worth it for the knowledge gained? Where do we draw the line? What is acceptable? As always, let’s begin with some questions: What does it mean to behave ethically? Central question: what makes a research study ethical or unethical? Who decides? ◼ Is it ethical to deceive a participant? ◼ Are there things that can’t be done under any circumstances, regardless of the benefit? ◼ How much does a participant need to know to provide consent to participate in a study? There is often a potential for risk for participants (either physical, mental or social) Deception may be involved Sensitive information is often collected Studying vulnerable populations (e.g., mental illness; children) presents unique challenges ◼ Are they capable of providing consent? ◼ Withholding treatment in a control group? Power differential between experimenter and participant ◼ Coercion is often unintentional Why the need for ethical rules? Some unfortunate history of abuses… Nuremberg Code (1947) ◼ Voluntary and informed consent is required ◼ Benefit must outweigh risk Declaration of Helsinki (1974) ◼ All research must be approved by an independent review board (IRB) ◼ Participant privacy must be respected National Research Act of 1974 ◼ Establishes guidelines for IRBs and ethical practices for research with human participants in the U.S. ◼ Under the auspices of the Department of Health and Human Services ◼ Protection of Research Participants Institutional Review Boards (IRB) Institutional Animal Care and Use Committees (IACUC) ALL research involving human or animal subjects must be reviewed prior to initiation ◼ Approve, Reject, or Modify ◼ Exempt Risk/Benefit Ratio Costs of doing or not doing the research ◼ A subjective judgment: Is it worth it? ◼ Do the benefits outweigh the risks? Risks: ◼ Physical Injury ◼ Social Injury ◼ Mental/Emotional Stress Benefit: ◼ To Participant ◼ To Science ◼ To People in the Future Minimal Risk The harm/discomfort of a study is not greater than would be expected in daily life or during routinephysical/psychological tests Naturalistic studies in public settings without identifying information ◼ Ex: Observation in a bar Normal educational/organizational practices ◼ As long as research won’t affect grades/employment status, and is confidential Minimizing Risk Responsibility to minimize risk in all cases  ◼ Some risk is acceptable, but should be minimized and outweighed by the benefits Use methods with the least risk possible ◼ Some methods can’t be used at all Anonymity Confidentiality and Privacy Protect Data Compensation must be non-coercive Note that there is an important difference between Minimal Risk and Minimizing Risk Informed Consent A PROCESS, not just a form ◼ Ensures participant: 1. Competence 2. Knowledge 3. Volition Must be documented, with a copy provided to the participant Steps to Informed Consent Explain purpose of research, procedures, time commitment Right to decline participation and stop at any time Potential consequences of stopping research mid-stream Potential risks​ Potential benefits Limits of confidentiality ◼ Harm to self ◼ Harm to others ◼ Child Abuse ◼ Abuse of Elders/Dependent Adults Incentives Contact information Answer questions Some participants can’t give consent Children and adults with mental disabilities IFC from parents or legal guardians  Should still get assent from participant For children >5 but