Exam 1 Review: Psychological Interventions PDF

xExam 1 review Psychological Interventions Week 1: Universal Interventions Universal interventions: are actions directed towards an entire population without regard for individual risk factors -​ Attempt to reduce risk factors and promote protective factors -​ Ex. D.A.R.E, free school breakf...

xExam 1 review Psychological Interventions Week 1: Universal Interventions Universal interventions: are actions directed towards an entire population without regard for individual risk factors -​ Attempt to reduce risk factors and promote protective factors -​ Ex. D.A.R.E, free school breakfast, universal pre-k Response to Intervention Selective Interventions Selective Interventions: are those that are delivered to a subset of individuals, families, or communities because their characteristics or exposures place them at risk for poor later outcomes Ex. head start for low-income families, daily aspirin for me over 50 Indicated: are directed towards individuals that are beginning to show substantial levels of difficulty, but usually are provided before the individual has received a formal diagnosis or special classification Ex. cholesterol lowering drug, individualized interventions in education Small Changes, Large Effects Adolescence Time of growth and change and risk -​ significant threats to the mental and physical well-being of today’s youth arise from risky decision making and dangerous activities Time when interventions often lose effectiveness (as adolescents get older) Real-World Contexts Basic research: investigations with limited clinical relevance that typically deal with theoretical concepts; conducted in highly-controlled lab settings -​ Ex. effect of growth mindset intervention that is given to college students in a controlled lab setting about changes in their beliefs about intelligence Applied Research: investigations in real-world contexts with more directly relevant results (more immediate to use) -​ Ex. effect of growth mindset intervention that is given to highschoolers about changes on their report card grades Approaches to Universal Intervention Skills-based approaches: change the PERSON -​ Social-Emotional Learning programs -​ DARE, abstinence only education LIMITS: -​ not solid evidence that targeted exercises to enhance skills actually improve life outcomes -​ Interventions might not alter the ‘right’ skills -​ Developing skills takes effort, practice, and repetition Situation-Based approaches: Change the SITUATION -​ ’nudge’ make organ donor ‘opt-out’ -​ opportunities ( age restrictions on alcohol consumption, pay-for-performance) LIMITS: -​ Change may not generalize for new situations -​ People may just not want to take advantage of new opportunities or resources -​ Some changes may be embarrassing Meaning-making Approaches (change person-in-context; Social Psychological Theory) -​ Focus on how individuals Interpret themselves and their circumstances Role of meaning-making in behavior Week 2: General Principle 1: Alter Meaning to Change Behavior 3 Core Motivations underlying meaning making ​ Need to understand, Need to belong, Need for self-integrity Need to understand: people want to develop understandings of themselves, other people, and the world, in order to predict their own or others’ behaviors -​ People act as ‘lay scientists’- trying to make sense of things that happen to us best as we can -​ Ex. if someone has a religious awakening, the way they will interpret that experience will be in light of their own religious background Need for self-integrity: -​ People want their interpretations to be based on reality, but we aren’t disinterested observers of the world. -​ People want to think well of themselves (want to believe they are adequate, coherent of their ideas)  -​ When this sense of self-integrity is threatened, we can become defensive and have poor outcomes as a result of that Stereotype threat: worrying about confirming the negative stereotype that exists about our groups; can threaten students integrity and lead to poor academic performance Need to belong: humans are inherently social species. People want to feel connected to others, be accepted and included, valued members of the group, and contribute positively to those groups. When this need is threatened, people often experience distress and dysfunction -​ ex. If a student who does not get into their favorite frat or sorority may feel that their need to belong is threatened (what’s wrong with me?) Alter the meaning: as psychologists we alter this negative meaning so the individual can still feel adequate, competent, and a valued individual General Principle 2: Meanings Operate Within Complex Systems 1.​ There are several influences on people’s interpretations of themselves and their circumstances 2.​ When we change people’s interpretations, only improve their outcomes when the necessary systems are already in place to lead them to a good path -​ Consequently, interventions may only work for certain people in certain circumstances (psychological precision), even though they might be delivered universally. 3.​ Likewise, interventions may work best when they are aligned with various other contextual affordances Ex. Draco Malfoy: Getting Draco to reinterpret his opinion of Harry may work best when having a high opinion of Harry would not carry enormous social costs (e.g., he would not take a huge social “hit”) Witherspoon Article Key Points -​ Gendered attrition: High performing women in pre-med courses experience significant attrition, Particularly in chemistry and physics, despite receiving good grades -​ Competency Beliefs: lower competency beliefs in chemistry is a critical factor influencing women’s decisions to leave pre med tracks -​ Motivational Factors: such as science identity and interest are important for keeping women in STEM fields -​ Need for Interventions: need targeted motivational interventions to support high-performing women in pre med courses, aiming to reduce attrition and promote gender equity in STEM education. Walton-Wise interventions Key Points ​ Wise interventions: social-psychological strategies designed to improve individual outcomes in various contexts, particularly in education. ​ Targeted strategies: self-affirmation and social belonging which address challenges faced by students, specifically from minority groups or disadvantaged groups ​ Stereotype Threat: impact of stereotype threat on performance and how wise interventions can mitigate its effects, allowing individuals to perform better in high-pressure situations. ​ Long-lasting Effects: The findings suggest that even brief interventions can lead to enduring changes in attitudes and behaviors, contributing to improved academic and social outcomes. LAB Real World application: Gender-based attrition from STEM majors Week 3: General Principle 3: Recursive Processes ​ Changes in people’s interpretations (meanings, stories) can be self sustaining and embedded in the structure of their lives ​ Recursive cycles can go from self-defeating to self-enhancing propelling gains forward in time or making them ‘snowball’--> pick up momentum and speed as it goes down the hill General Principle 4: “Wise” Techniques -​ Wise interventions thus use strategies that have been honed through basic research to offer people constructive meanings in compelling ways -​ Less is more in these situations, people often don’t even see themselves as having a problem that needs solving (people may be resistant) ​ Direct labeling: ○​ providing people with a positive label that defines an otherwise ambiguous situation (e.g., “You’re a “hard worker;” You’re a “good helper”) ​ Prompting new meanings:  ○​ give people basis for making new interpretation; prompt them to reconsider how they think about the world (provide new facts) ​ Increasing commitment through action: ○​ People want to see their behaviors and attitudes as consistent ○​ Creating situations that inspire people to act in a way in line with a new idea (ex. having someone advocate for recycling, will lead them to recycle themselves) ○​ ‘Do good be good’: by acting in a way that we think is a good thing, leads us to interpret ourselves as being good ​ Active reflection exercises: ○​ A few exercises that help people reframe their experiences, often through writing ○​ Not giving people new info and not asked to change behavior, but rather asked to reflect in a different way on a challenging/positive event in their lives ​ Ecological Belonging Intervention in STEM courses ○​ Gendered attrition- women are leaving stem/premed fields more often than men ​ Problem because it can restrict career opportunities, slow economic growth, less viewpoint diversity and fewer breakthroughs,less optimal care, etc. The Ecological-Belonging Intervention: Binning reading Main components: ​ Collective norms: importance of having collective norms within the classroom frame adversity as a common and temporary experience. This helps students understand that challenges are shared and can be overcome ​ Social ecology: includes interactions and relationships among students and between teachers and students. Modifying this ecology, the model can create a supportive environment that fosters belonging ​ Long-term effects: interventions based on this model can have lasting impacts on students’ trajectories, such as higher GPAs and college excellence Lab Activity 3 Week 4: Developmental period of adolescence Adolescence begins with the onset of puberty and ends with a transition to an adult role in society Middle Adolescence: period after pubertal maturation but before transition to an adult role in society (grades 7-11) ​ Tremendous growth (exploration and learning) + logical and moral reasoning and heightened risk for mental and behavioral health problems ○​ Reason significant threats to the mental and physical well-being of today’s youth arise from dangerous activities in which they willingly engage. Hypotheses for Understanding and Improving Adolescent Interventions ​ Hypothesis 1: Desires for status (S) and respect ® ○​ Status: one’s place in social hierarchy ○​ Desire: the desire to have high social rank ○​ Respect: being granted rights one expects to be granted in one’s role in society  ​ Feel respected and that they have high status when treated as competent, have autonomy, and bring value to their desired group Sensitivity to status and respect: prepared to align their attention, motivation, and behavior with rewarding feelings of having status and being respected ​ Hypothesis 2: Traditional Interventions become less effective during adolescence because they do not honor desires for S and R, both in what they say and how they say it What they say: ○​ Focus on providing knowledge or skills with the goal of preventing short-term desires for the sake of long-term goals. ​ Ex. D.A.R.E → No to drugs right now, so have health in future ○​ Traditional interventions also fail to deal with main reasons for engaging in behavior (when they are so focused on long-term goals) How they say it: ​ Focus on lectures, assemblies, homework– these methods can backfire even if effective at communicating complex info ○​ These lectures can feel disrespectful, since they already know this info (feels like nagging) ​ If these intervention messages are perceived as uncool or low status, it could be disregarded by the adolescents (they don't want to seem low status) ​ Hypothesis 3: Improved interventions could honor the sensitivity to status and respect and thereby capture adolescent attention and motivation to create behavior change Yeager et al. Reading ○​ Aligning Interventions with Adolescent Needs: creating environments where adolescents feel valued and heard, providing ○​ Examples of Successful Interventions: programs that frame healthy behaviors as desirable and programs that allow adolescents to make decisions can increase respect over their own actions ○​ Understanding Sensitivity: Adolescents are particularly attuned to feelings of status and respect, which can significantly influence their behavior. When interventions fail to acknowledge this sensitivity, they may become ineffective. ​ Comparing abstinence education vs. 16 and Pregnant on pregnancy and birth rates LAB 4 Reading: Trenholm- Impacts of abstinence education ​ Found that abstinence-only programs did not impact onset of sexual activity among teens ​ These programs may enhance knowledge of STDs, but do not significantly influence teens' sexual behaviors or reduce the risks of pregnancy and STDs Week 5: Measurement ​ Construct: what are we trying to measure? What is it? ○​ Ex. ‘teen pregnancy’, ‘self-efficacy’, ‘intelligence’, ‘school readiness’ ​ Operational definition: description of a construct in terms of procedures, actions, or processes which it could be observed and measured ○​ defining the construct in a way that it is measurable ​ Ex. teen pregnancy can be operationally defined as number of pregnancies a person has from ages 13 to 19 ​ Types of measures (questionnaires, performance measures, administrative records) ○​ Self report questionnaire: most common approaches to assessing personal qualities among both researchers and practitioners ​ Cheap, quick, reliable, and predictive of outcomes ○​ Objective instruments (Performance measures): designed to elicit meaningful differences in a behavior of a specific kind. (no subjective interpretation) ​ Academic Diligence Test (ADT) ​ Students allocate time between math problems or playing video games (measuring self-control through behavior) ​ More time student spends solving math problems reflects self control (more time = higher self control) ​ Marshmallow Test ○​ Administrative Data: we can measure a construct by passive instrumentation (school records, voting records, arrest records, etc.) ​ Ex. Teen Pregnancy– search hospital records or publicly available data on birthrates hat would correspond to when the subject was between the ages of 13 and 19 ​ Reliability and Validity ○​ Reliability: whether an assessment tool produces stable and consistent results ​ Ex. You measure a child’s self-control multiple times under the same conditions. The self-control instrument displays same score everytime, so the test is reliable ○​ Validity: How well a test measures the ‘construct’ it is supposed to measure ​ Does the self-control measure actually measure self-control (and not something else)? Target practice ○​ Red dot in the middle is construct we are measuring 1: reliable because same scores are being produced, but not valid because it isn’t hitting the bullseye 2: valid because the dots center around the bullseye, but not reliable because we don’t know if its measuring the right thing 3: neither reliable or valid because it’s producing scattered results (reliable) and not hitting bullseye (valid) 4: both reliable and valid because produces consistent scores and hitting bullseye each time Factors that Affect Reliability of Questionnaire and Performance Methods ​ Intrinsic factors: have to do with the instrument itself ○​ Length of instrument, consistency of items, item difficulty, instructions, training of the scorer, subjectivity of the score, calibration errors ​ Extrinsic factors: remain outside of the instrument itself ○​ People's mood changes on a moment to moment basis, effects consistency of information instrument is able to provide ​ Threats to reliability and validity across measurement type ○​ Factors that Affect Validity of Questionnaire Measures: misinterpretation by participant, lack of insight or info, reference bias: implicit standards used when making judgments may differ across individuals, faking and social desirability bias, ○​ Factors that Affect Validity of Performance Measures: misinterpretation of researcher (make assumptions about underlying reasons for individual’s behavior), artificial situations, task impurity (Task performance may be influenced by irrelevant competencies (e.g., hand-eye coordination) ○​ Factors that Affect Validity of Administrative Data: incomplete coverage of construct: even though this data may measure the construct well, it might not measure the entire spectrum of construct ​ Ex. using cafeteria purchase records to tell us whether high schoolers eat healthy could tell us their behavior at school but cannot tell us what they eat outside of school ​ Lab Activity on comparing measurement types Duckworth article key points  ​ Self-report and teacher-report questionnaires have limitations like reference bias, lack of insight, and insensitivity to short-term changes ​ Measuring non-cognitive attributes such as grit,self-control, and emotional intelligence is very important, but there are challenges in doing so Week 6: Randomized Experiments ​ 3 conditions necessary for establishing causality ○​ Cause and effect must be empirically related (correlation) have to be correlated! ​ Ex. smoking and lung cancer must be correlated ○​ Cause must occur before the effect (temporal precedence)--> not sufficient for ruling out other explanations ​ Smoking must come before lung cancer ○​ control for all third-variable confounds (rule out all alternative explanations) ​ Relationship between smoking and lung cancer is NOT EXPLAINED by genetic risk, family factors,cultural factors Why Doesn’t Correlation Imply Causality? 1.​ Correlation might just be a coincidence (ex. correlation between age of miss america contestants and murders by steam objects) 2.​ Reverse causality is a possibility→ when the effect of something actually causes the thing itself (ex. Because politicians don’t do what I want, I don’t vote) 3.​ Other factors may explain relationships (i.e., confounding variables)--> (ex. A third variable (summer time) may cause us to link to things together (forest fires and ice cream sales) ​ Comparing study designs (pre-post, quasi-experiment, true experiment) ​ Pre-Post design: Pre test/ Post Test design: people’s attitudes, beliefs, and behaviors before intervention are compared to attitudes, beliefs, and behaviors after intervention Ex. Let’s say I get 100 people from Company X to volunteer for my 16-week smoking cessation program. -​ Before the program, all 100 people were heavy smokers. -​ After the program, 68 people are smokers and 32 people are nonsmokers Does my intervention work or not? ​ Good news is that these studies are relatively cheap to conduct... ​ But there are serious problems with this design in terms of causal inference. Big Problem: No Control Condition!  -​ Without a comparison condition (i.e., people who do not receive the program), we cannot meet the necessary requirements of causal inference. Why? -​ Because everyone gets the program, we cannot compute a correlation between our supposed cause (smoking cessation program) and the outcome (quitting smoking). -​ A correlation requires variation in both the cause and the effect. We have no variation in the cause in pre-post designs. Another example: If all study participants are men, we cannot determine whether a treatment works better for men vs. women (no variation in gender among our participants) -​ Because we cannot show a correlation, we also cannot rule out the possibility that things other than the intervention led people to quit smoking Threats to causal inference (history, maturation, statistical regression, selection bias) 1.​ History Confound: Events occurring during the experiment that are outside the researchers’ control (e.g., launch of a national anti-smoking campaign, global pandemic shuts down stores) 2.​ Maturation Confound: Natural changes that occur with passage of time (e.g., people had planned to quit anyway) 3.​ Statistical Regression Confound: Phenomenon where most extreme values in a population eventually move closer to the mean (e.g., by selecting a group of heavy smokers, some might eventually rebound and quit) 4.​ Countless other confounding factors might have led to a reduction in smoking during the 16-week period that had nothing to do with the intervention (e.g., cancer diagnosis, family pressure to quit) Cannot establish temporality and cannot rule out alternative explanations! ​ Lab comparing control vs. intervention over time (condition-by-time interactions) (2) Quasi-Experimental Study: Study design in which we compare attitudes, beliefs, and behaviors among people who participate in an intervention to those who do not participate. Ex. Now, let’s say I get another 100 volunteers from Company X to take my 16-week smoking cessation program. -​ For comparison purposes, I will get 100 volunteers from Company Y but they won’t receive the program. -​ I will compare rates of smoking before and after the 16-week program for people who participate vs. those who do not Good news is that I’m able to show a correlation between my smoking cessation program and reductions in smoking -​ Those who get program quit at higher rates than those who do not get it (i.e., program enrollment is correlated with reductions in smoking) -​ We can also rule out History, Maturation, and Statistical Regression Confounds. We know this because the comparison group presumably experiences the same history and  maturation as the experimental group, allowing us to know the reduction in smoking rate that would have happened anyway without a program In other words, we can also establish temporal precedence, since people enroll in the program before the reduction in smoking occurs (not the other way around) Problem 1: Selection Bias! -​ Without randomly assigning people to smoking cessation program vs. no program (control condition), we still cannot rule out the possibility that things other than the program might have influenced people to quit smoking. -​ Differences in the outcome (smoking) between groups could be attributed to initial differences rather than a treatment effect -​ People at Company X may have volunteered to participate because they already wanted to quit smoking, whereas people at Company Y might not have cared about quitting. Thus, people who are motivated to quit might be motivated to enroll in a program. (3) Randomized (Controlled) Experiments: Type of study in which people are divided into groups randomly (e.g., based on coin flip). People have an equally likely chance of being in the treatment group or the control group. -​ Random assignment is the Great Equalizer! -​ By randomly assigning people to groups, we can be confident that they do not differ in their family history of cancer, motivation to quit smoking, personalities, health care plans, hobbies, or in any other way that might influence their behavior. -​ It would be extremely unlikely that one group would have significantly more cat lovers, vegetarians, hockey fans, or marathon runners than the other, just as it would be extremely unlikely for a fair coin to turn up heads forty times on fifty flips. With random assignment, we rule out all other factors that might confound the relationship between the supposed cause (smoking cessation program) and outcome (smoking reduction) Now, let’s say I randomly get 200 volunteers from Company X and Y to take my 16-week smoking cessation program. -​ Everyone is randomly assigned (by a coin flip) to receive either the smoking cessation program or nothing. -​ I will compare rates of smoking before and after the 16-week program for people who are randomly assigned to receive the program vs. those who do not Satisfies all 3 requirements for causal inference: -​ Shows correlation between supposed cause (e.g., smoking cessation program) and outcome (e.g., smoking reduction) -​ Shows temporal precedence because supposed cause (e.g., smoking cessation program) occurs before the outcome (e.g., smoking reduction) -​ Rules out all alternative factors that might confound (account for) association between supposed cause (e.g., smoking cessation program) and outcome (e.g., smoking reduction) Two caveats: -​ Quality of control condition: Inferences regarding treatment effectiveness are always made in comparison to another group. If we have a weak control condition (e.g., no- treatment control), then we cannot make as strong an inference about treatment than if we had strong control condition (e.g., comparison to gold-standard smoking cessation program) -​ Quality of measures: Inferences regarding treatment effectiveness also depend on type and quality of measures used! Lab 6 Statistical Information ​ P-values: measure of the probability that an observed difference could have occurred just by chance. Can be used when comparing means of two or more groups, when examining correlation or regression coefficients, and so forth. We use p-values to signify whether the comparisons or coefficients are statistically significant. ​ Conventions: ○​ P-values lower than.05 are considered statistically significant. (unlikely to be due to chance) ○​ P-values greater than.05 are considered not statistically significant. (cannot rule out possibility that results are due to chance) ​ Example: ​ The height for adolescent boys is M = 69.6 inches and for girls is M = 64.8 inches (p =.003). ​ Are the group means significantly different or not? Question Types ​ True/False ​ Multiple Choice (single answer) ​ Multiple Answer (select all that apply) ​ Short Answer: Open-ended questions that will require up to about a paragraph of writing