PSYCH 257 - Chapter 4: Research Methods

hypothesis

educated guess on what we expect to find from the study

prospective studies

record changes over time as they occur

ex. ASD studies that follow kid with genetic vulnerability to autism, see that the changes happened gradually not suddenly.

often find that whats actually happening in development of disorders or treatment is quite different from what people remember.

health promotion/positive development strategies

efforts to help entire populations of people by preventing development of the disorder through interventions

ex. skill building for kids in high-crime rates so they can learn better and engage in positive behaviours. long-term follow up suggests multiple positive effects in achievement… later in age

universal prevention strategies

focus on whole population but target certain risk factosr (ex. behaviour problems in inner-city classrooms) without focusing on specific indivudals.

specific risk factor, not people.

selective prevention

targeting groups at risk (people now) within a population, designing specific interventions so they dont get into trouble later on!

indicated prevention

strategy for indiviudals who are beginning to show signs of problems, but dont have a psychological disorder yet

ex. ealry signs of depression, but no disorder yet. interventions focus on preventing mild symptoms from further developing through getting them a support system and tackling the unhealthy thoughts quickly.

research design

the design for the study you are using to find out if ur hypothesis is correct or not

independent variable

the variable that you are manipulating.

ex. in a study on the relationship between panic attacks and alcohol use, the independent variable is alcohol use, because the amount of alcohol someone consumes can be manipulated

dependent variable

the effect being measured in a study. its value is DEPENDENT on the changes made by the researcher.

ex. in a study on relationship between panic attacks and alcohol use, the dependent variable is the intensity of panic attacks (of which value changes depending on the degree of alcohol consumption)

internal validity

the extent to which we can be confident about the establishment of a cause-and-effect relationship without the existence of confounding variables.

the extent to which results can be attributed to the independent variable

these are highest in experiments because of the controlled environment! it is lowest in correlational studies because there is zero manipulation and control for confounding variables.

factors that affect internal validity

1. confounding variables (ex. drownings increase with ice cream sales, confounding variable: weather)

2. selection bias (non-random assignment, volunteer bias, attrition bias)

3. cohort effect (differences between groups in study are due to unique experiences of that cohort, meaning the findings cannot be generalizable to people in general (ex. COVID and gen z/gen alpha, affecting the cohort in ways are more sensitive to PTSD for example, and this isnt because of age!! its because of the specific things the generation dealt with)

4. testing effects (changes in partiicpants behaviour due to repeated exposure to the same test because humans can learn and get better at things, or we get bored!!! ex. redoing the same puzzle over and over again, you get better at the puzzle cause you learned how to do it, or you get really bored.

external validity

extent to which results can be generalized outside the immediate study

usually lower in experiments because it might not show how it goes in real life. usually higher in correlational studies cause its looking at whats acutally happening in real life. ex. experiment of overprotective parents on social anxiety in kids is gonna have low external validity cause it doesnt really tell you how overprotective parents affect children in real life.

testability

the ability to either accept or reject a hypothesis.

ex. astrology is not testable cause how tf are u supposed to know if a star is affecting everyones behaviour

confounding variable

any factor in a study that makes the results uninterpretable because a variable other than the independent variable affects the dependent variable (ex. something other than type of treatment affects scores on anxiety scales)

control group

exact same type of people as experimental group but they are exposed to a placebo instead of treatment (or TAU - treatment as usual)

randomization

assigning people to different groups COMPLETELY RANDOMLY which means everyone had equal chancd of being placed in any group (ex. flip coin, assign everyone a number, pick every 7th person)

analogue models

controlled conditions of lab that are comparable (analogous) to phenomenon under study

ex. binge eating in the experiment, questioning before, during, and after binge eating. instead of just a scale.

ex. phobias: simulating spiders in a room for someone with a spider phobia to figure out how they ACUTALLY react to spiders

ex. stroop task for anxiety: colours and names of colours are swapped, which stresses the person out!! and simulates how they would respond in real life!!!

generalizability

extent to which results apply to everyone with a particular disorder

statistical significance

mathematical calculation about difference between groups (is there a statistically significant difference between the two treatment methods used to treat anxiety?)

clinical significance

even if the difference is statistically significant, is the difference actually MEANINGFUL in the LIVES of those who are affected with a disorder

ex. two different types of medication. one medication helped people for 75 days, the other for 77 days. this was statistically significiant BUT its not a significant difference in the lives of the people with the disorder, so it is not clinically significant.

effect size

tries to find how big the difference is between the two groups/variables.

ex. studying exercise, improves mood, if mean mood score of people who exercised was 10 points higher than those who didn’t, in relation to the standard deviation, u can find the effect size, if the d-value is 0.67, the effect size is medium large, and so its meaningful.

social validity

involves getting input from the person being treated, and people who are close with them, about the importance of the changes that have occured.

if the effect of the treatment is large enough to impress those who are directly involved, treatment effect is clinically significant.

patient uniformity myth

comparing groups based on mean scores hides important differences in individual reactions to interventions. we cant assume everyone will react the same to a treatment, even if the findings are statistically significant.

ex. group A improved by 50% over group B

ex. 2 people with alcohol use disorder. 1 will thrive under treatment, which is in accordance with the data, but the other might not! and while the study says a certain thing, for the person whose life is actually on the line, they aren’t thriving!! and we’re in this to make them thrive!!

case study method

1 person or 1 group of people with disorder. HEAVYYY FOCUS on this person or group of people to undresatnd like everything about them to learn more about the disorder.

ex. sexual attraction to horses, extensive case study on that person to figure out how and why they got here.

adoption studies

we have an adoptee who has disorder, find his brother who’s in a different home, try to see if he displays signs of disorder too. if they can identify enough sibling pairs, they can find out whetehr siblings broght up in different families display the disorder tot eh same extent as the proband (the first subject with the disorder) to learn about the role of genetics in the disorder.

association studies

compare the genetic markers in two different large groups:

group 1: has disorder

group 2: doesn’t have disorder

they look for genetic idfferences between these two groups. if group 1 has marker X showing up a lot, and not as much in group 2, there is an association between marker X and schizophrenia.

only associaotn by the way! not cause-and-effect

baseline

condition is evaluated before treatment! ex. anxiety levels before drugs are administered.

another way to think about it is taking the baseline of someone’s memory before a brain surgery to figure out if memory deteriorated after surgery.

cohort effect

confounding variable of age and variable

ex. gen z is more anxious than millenials, is this cause of age? nah its cause of COVID man. our collective experience changed the way we think, feel and act and that cant be attributed to age.

cohorts

the participants in each age group/generation

ex. study of 12, 15, 17' year olds, each of those age groups are a cohort

comparative treatment research

gives different treatments to two or more comparable groups of people with the same disorder, to figure out which one yields better outcomes for them.

ex. CBT vs medication (the 2 different groups) —> for depression (the same disorder). figure out what works better!

correlation

statistical relationship between two variables

ex. correlation between schizophrenia and size of ventricles in the brain

correlation coefficient

vary from -1.0 (perfect negative correlation) to 1.0 (perfect positive correlation)

cross-generational effect

the way experiences and values are passed down from one generation to the next, shaping beliefs and behaviours across generations

ex. great depression meant ur parents grew up poor, that affected how they viewed money, now you are also careful with money

this is important to know because it helps us see how family histories and experiences shape how groups of people think

cross-sectional design

(age version):

• researchers take cross section of pop across different age groups and compare them on a characteristic

• ex. ages 12, 15, 17 interviewed on beliefs on alcohol use to provide a “snapshot in time” on what they think

dependent variable

double-blind control

both experimenters and participants have no idea whose in which group. prevents experimental biases in the way they treat patients, prevents participants getting better cause they wanted to, gettting wrose cause they were disappointed.

endophenotypes

genetic mechanisms that contribute to underlying problems causing symptoms experienced by ppl with disorders

ex. theres no schizophrenia gene! theres genes responsible for the working memory problems schizophrenia is characterized by

epidemiology

the study of incidence, distribution and consequences of a particular problem in a population

ex. incidence, prevalence…

experiment

manipulation of independent variable and observation of its effects

in a controlled environment often

family studies

examination of a behavioural pattern or emotinal trait in context of family

used to see prevalence in the family tho genetically, but the issue is that they live together so shared environment can affect in socail adn psychological aptterns (ex. modelled behaviour thorugh childhood influencing them)

genetic linkage analysis

some genes are close together on a chromosome, and since theyre so closed together, they tend to be passed down together.

in these studies, we look at the dna of a bunch of family members and try to find genetic markers that are passed down along with schizophrenia (ex. marker X is always inherited by family members with schizophrenia, but not those who don’t).

from this, we can figure out that the gene responsible for schizophrenia must be located near marker x, and from this can narrow down where the gene that infleunces schizophrenia is located

genetic markers

characteristics that we know the location of during family disorder studies.

ex. SNPs —> people with schizophrenia always have a G wehre people wtihout schizophrenia always have an A. this is a genetic marker. it helps us get closer to figure out where the gene that plays a role in schizophrenia is located.

genotypes

genetic makeup (Bb eyes)

human genome project

mapping of 20k human genes, identifying genes that contribute to inherited diseases

incidence

estimated number of NEW cases during a specific PERIOD.

there is a TIME PERIOD. think “incidentally” in reference to something just happening in a spot of time.

informed consent

participants formal agreement to cooperate in the study following full disclosure of the nature of the research and the participants role in it.

1. competence

2. voluntarism

3. full information

4. comprehension

longitudinal designs

follow a group over time; used to assess indivdual change

ex. study on physical aggression in boys, following 1000 boys from low-income neighbourhoods from age 6 to age 15.

multiple baseline

researcher starts treatment at different times across settings, behaviours, or people.

ex. wenday’s anxiety is measured at both (1) home and (2) work. it isolates the effects of the intervention across different environments.

1. baseline (measured repeatedly at both home and work WITHOUT TREATMENT). wendy may have more anxiety at work than home, so its important to know.

2. sequential intervention:

   1. (1) treatment introduced at home, if anxiety reduces only at home but not at work, then treatment is effective AT HOME

   2. (2) treamtnet is then introduced at work to see if similar improvements work there

      If Wendy’s anxiety improves only in settings where treatmnet is applied, seems like treatment is responsible for improvement! cause confounding variables are being taken care of!!

   3. internal validity is enhanced by removing confounding variables.

negative correlation

relationship between 2 variables is reversed. they are inversely correlated. (ex. greater social connection, less anxiety)

phenotypes

observable characteristics (blue eyes)

placebo control groups

placebo is given to control group to make them beleive they are given treatmnet. can be a pill, or a psychological treatment as usual (TAU) without the element that is being measured.

placebo effect

when behaviour changes as a result of a person’s expectation of change instead of because of what the experimenter did

positive correlation

higher scores in variable is related to higher scores in the other variable (more maritial distress, more disruptive behaviour in child) and vice versa (but they are moving in the same direction)

prevalence

the number of people with a disorder at any one time. TIME PERIOD HERE TOO.

ex. 7% of U.S. population in any given year.

proband

the FAMILY MEMBER iwth the trait singled out for study. it is the FAMILY MEMBER not the trait.

repeated measurement

used in SCED. behaviour is measured several times instead of just once before you change the independent variable.

1. baseline phase (A): data collected before independent variable to establish a baseline

2. intervention phase (B): repeated measurements continue to look for changes that occur as a result of treatment

3. withdrawal phase: if intervention is withdrawn, we can tell if treatment acc did anything lol

retrospective information

ifnormation gathered as people are looking back

ex. asking adults if they were anxious about school when they were children

its usually less accurate because memories are falliable. u might remember yourself being way more or way less anxious tahn you actually were.

sequential design

longitudinal + cross-sectional

repeated study of different cohorts over time

ex. development of alcohol and drug us among BC youth, data from 1300 students in school district from grades 7, 8, 9 (cross-sectional) + over 3 years (longitudinal)

single-case experimental designs

research methods used to study the effect of a treatment on a single indivudal or small group by observing changes in behaviour over time. they are literally experimental designs, using a SINGLE-CASE.

like case studies, but they use strategies to improve interna validiyt cause they are experiments.

treatment outcome research

trend

general movement (line that goes up or down) to figure out if treatment is helping or not.

twin studies

do twins share the same trait? we can find this for physical traits pretty easily, but to find it for psychologcail traits, you combine adotion studies and twin studies, where u find twins that were seperated at birth and see if they have the same disordre to find out mroe about role of genes and environment in development of behaviour al patternsa nd symptoms of disorders and stuff

variability

fluctuation in the data collected, possibly from day to day. looking for a general trend. variability can be imortant in interpreting effect of treatment tho, if theres variability that shows that she was fluctuating around a lot, or decreasing on her own, maybe the treatmnet wasnt the thing that actually helped!

withdrawal design

researcher tries to figure out whether IV is responsibile for changes in behaviour or not. so they do a baselilne, then the treatment, then removes the treatmnet to see improvement from after baseline, then deterioration after treatment is withdrawn.

issues with case studies!!

1. things about the person you’re stuyding may be just unique to them not generalizable to the entire public

2. ex. guy who’s attracted to horses had low IQ, said those with low IQ can be linked to bestiality and zoophilia, but another case study disproved this!

issues with correlational studies

1. HARD TO FIND DIRECTIONALITY!!! are socially anxious children creating overprotective parents, or are overprotective parents creating socially anxious children? if in a correlational study yk

clinical trial

evaluation method that follows number of generally accepted rules used to determine the effectiveness and safety of a treatment

controlled clinical trials

used for experiments that have a control group to figure out which treatment method is better. used for comparison! as there is a control and experiment!

treatment outcome research

systematic study and evaluation of effectiveness of treatments. basically study of treatments man.

resentful demoralization

may be disappointed theyre not getting treatment and score even worse than they wouldve if they didnt know

placebo control group

used to address placebo effect

the control group is given a pill that looks very similar, but it doesnt have the medication in it.

for psychological treatments, one way they do it is by taking a worksheet that everyone gets, but the one thing that they think is going to actually help is only on the experimental group sheet.

allegiance effect

if the treamtnet that was supposed to work seems to be failing (and researcher would know whose in what) they may not push as hard!!

process research

research on the mechanisms responsible for behaviour change. asking why something works. it is important to find the “active ingredients” which are ACTUALLY what are causing a change, so we don’t waste time and money on things that don’t actually do anyhting.

functional communication training

1. use of multiple baseline: introducing treatment to group of 5 children

2. DV: incidence of behaviour problems + new comm skills

3. only once treamtn was adminstered did change happen

4. this multiple baseline thing let us rule out confoudning variables.