Final Review Day 3

Front 1

Logistic Regression - Overview

Back 1

Used to measure PROBABILITY (0-1) / BINARY / U-SHAPED, NOT NORMAL DIST

1) A technique for fitting a regression surface to data incl a dichotomous dependent variable

2) often preferred instead of linear regression

3) can be used to predict group membership from a set of variables

4) can be used to provide knowledge of the relationships and strengths among the variables

5) differs from Multiple Regression in 3 ways

~logistic coefficients are partial coefficients, controlling for other variables in the model, whereas Correlation coefficients are uncontrolled

~logistic coefficients reflect linear and nonlinear, whereas correlation reflects only linear rels

~a significant logistic reg means that there is a rel btw the ind variable to the dependent variable for selected control group but not necessarily overall

Front 2

Logistic Regression - Binary Logistic Regression Analysis

Back 2

1) the closer the odds ratio (OR) value is to one, the more likely it is that independent variable categories will not be found to significantly infl other factors

2) the closer the odds ratio (OR) value is to one, the more likely such factors will be found to be independent of the dependent variable

3) a value of one represents full statistical independence

4) the OR is a measure of effect size

Front 3

Stepwise Multiple Regression - Overview

Back 3

1) also known as statistical regression

2) involves adding or subtracting predictors one at a time and calculating the multiple correlation coefficient to use as a predictive power

3) Used:

~in exploratory phase of research

~for purposes of pure prediction

~not for theory testing

4) the ultimate goal is to identify the fewest # of predictors needed to account for the largest amount of variability in the study

5) consists of forward selection analysis and backward elimination analysis

Front 4

Stepwise Multiple Regression - Steps for Forward Selection

Back 4

1) add the independent variable (IV) with the highest significant correlation with the dependent variable (DV)

2) add the next IV w/ the highest significant semi-partial correlation with the DV

3) the regression equation is then recalculated

Front 5

Stepwise Multiple Regression - Steps for Backward Elimination

Back 5

1) Enter all IVs into the equation

2) delete the IV w/ the lowest non-significant semi-partial correlation w/ the DV

3) the regression equation is then recalculated

Front 6

Stepwise Multiple Regression - drawbacks

Back 6

1) May cause generalizations across datasets (unreliable)

2) r-squared estimates that are substantially too high

3) significance tests which are too lenient (increasing Type I Error)

4) confidence intervals that are too narrow

Front 7

Variables

Back 7

1) unlike constants, which remain the same, variables can take on many values

2) IV, or predictor variable: manipulated by the researcher, presumed to be the agent of change (i.e. affects the DV)

3) DV, or criterion variable: measured by researcher to determine if IV has an effect

4) Confounding variable: extraneous variable that varies systematically with the IV, thereby reducing Internal Validity (i.e. the ability of the researcher to claim differences in DV are due to the IV)

5) Quasi-independent variable: IV in quasi-experiment (i.e. experiment using existing groups rather than Random Assignment in determining condition)

6) Measurement scales:

~Nominal variable: a label or catagory (e.g. political party)

~Ordinal variable: data are ranked, or possess order (e.g. class rank)

~Interval variable: ranked, meaningful differences btw values (e.g. temperature scale, F or C)

~Ratio variable: ranked, meaningful diff btw values, and the value of zero signifies absence of what is measured (e.g. Kelvin temp scale) - TRUE ZERO

Front 8

Validity - Construct

Back 8

1) Convergent: a test is correlated w/ other tests measuring the same or similar trait

2) Divergent: a test is NOT correlated w/ tests measuring unrelated traits

3) Factorial: using factor analysis to show validity; factors that should correlate do, and factors that should not correlate do not

Front 9

Validity - Criterion-Related

Back 9

1) Concurrent: a test is correlated w/ a criterion variable

2) Predictive: a test is correlated w/ a criterion variable measured at a future time (e.g. SAT score and college GPA)

Front 10

Validity - Content

Back 10

Adequacy of test in measuring all facets of a construct or trait

Front 11

Validity - Face

Back 11

1) Assessed by those taking the test; the extent to which a test seems to measure what it it intended to measure

2) A valid measure is reliable, but a reliable measure isn't necessarily valid (e.g. a broken watch)

Front 12

External Validity - Def, 5 main threats

Back 12

Definition - the extent to which research findings may be extended to other people, places, and situations

5 Main Threats

1) Interactions btw diff treatments: the interaction effect btw factors is diff from the main effects

~Exp: a new instructional method promotes greater learning, but only among those who receive a motivational incentive

~examine interaction and main effects

2) Interaction btw testing and treatment: measurement sensitizes participants to, or inoculates them against, the tx

~include a post-test only control condition

3) Interaction btw selection and tx: diff effects of a tx for diff types of ppl

~exp: an experimental drug works for women, but not for men

~obtain as heterogeneous a sample as poss

4) interaction btw setting & tx: diff effects of a tx in diff settings (e.g. classroom vs combat zone)

~perform study in all relevant settings

5) Interaction btw history and tx: diff effects of a tx at diff times (e.g. pre vs post 9/11)

~replicate study at diff times

~perform lit review on earlier findings

6) Reactivity: changes in beh as a result of a person being observed

7) Demand Characteristics: when a participant behaves accd to what they think is expected

8) Carryover effects: occur when the effects of one tx are carried over to the next

9) Sequence or order effects: when the order of tx in a series infl the participant's response

~counterbalancing: a control measure against order effects conducted by presenting tx in a variety of sequences; w/in subject and w/in group

Front 13

Internal Validity - Def, Supports

Back 13

Definition

1) the extent to which a research study rules out alternative explanations and establishes causality

2) If not seeking to estab that A causes B, then Internal Validity is not a concern

3) If seeking to estab that A causes B, then Internal Validity is of paramount importance (more of a concern than External Validity)

Supports (ways to increase)

1) Random assignment is the most effective method for obtaining Internal Validity

2) Matching: participants are matched across conditions on a characteristic related to the outcome variable

3) Blocking: a characteristic related to the outcome variable is incorporated into the design as an additional IV

Front 14

Internal Validity - Threats

Back 14

1) History: An extraneous event, outside of the study, occurs during the study period

2) Maturation: respondents change systematically over the course of a study

3) Testing: Initial measurement affects subsequent measurements

4) Instrumentation: changes in the measure may skew results

5) Regression to the mean: extreme scores tend to be less extreme on subsequent measurements

6) Selection: participants differ systematically by experimental group b/f the intervention

7) Mortality (or dropouts): diff dropout rates across conditions render experimental groups non-equiv

8) interaction w/ selection: several of the above threats may interact w/ selection and be mistaken for tx effects

9) ambiguity about the direction of causation: non-experimental studies cannot confirm whether A causes B, B causes A, or a third variable C causes both A and B

Front 15

Internal Validity - Ways to Control Extraneous Variables

Back 15

1) Elimination: Complete removal of variable

2) Constancy: variable experienced by all participants

3) Balancing: matched-pairs design used to evenly distribute variables across groups.

Front 16

Reliability - Def, Methods for Determining

Back 16

Definition

1) Extent to which a measure or test is consistent and repeatable

2) necessary, but not sufficient, for validity

Methods for Determining

1) Internal consistency: all items measure the same thing

~reliability coefficient

~split-half

~Cronbach's alpha (coefficient alpha)

~KR 20

2) Consistency btw alternative forms

~coefficient of equivalence

3) Test-retest consistency

~coefficient of stability

Front 17

Reliability - Other forms of reliability

Back 17

1) Inter-rater reliability

~Kappa coefficient is used w/ nominal or ordinal data

Front 18

Measures of Central Tendency - Central Tendency, Mode

Back 18

Central Tendency - where the bulk of the distribution is centered

Mode

1) the most freq occurring value

2) distribution may feature one, two, or more modes

3) Advantages:

~applicable to all measurement scales (e.g. nominal, ratio)

~unaffected by extreme values (outliers)

4) Disadvantage: does not lend itself to algebraic manipulation (i.e. being placed in equations)

Front 19

Measures of Central Tendency - Median, Mean

Back 19

Median

1) middle value

2) Advantages:

~applicable to all but nominal data

~resistant to extreme values/outliers

3) Disadvantage: Does not lend itself to algebraic manipulation

Mean

1) average of all values

2) Advantages:

~lends itself to algebraic manipulation

~more stable estimate of central tendency (i.e. less sample to sample variation)

3) Disadvantages:

~requires interval or ratio variable (i.e. needs meaningful diffs btw values)

~affected by extreme values/outlier

Front 20

Measures of Central Tendency - Cross-distribution comparisons

Back 20

1) Normal distribution: Mode = median = mean

2) Positively skewed: Mode < median < mean

3) Negatively skewed: Mean < median < mode

Front 21

Constants

Back 21

1) a constant is fixed, unvarying value (e.g. 5)

2) when a constant is added to or subtracted from a variable, the results are as follows:

~measures of central tendency (e.g. median or mean) change similarly; adding 5 to every value results in a mean of 5 more than the original

~measures of variability (e.g. range or standard deviation) remain the same; adding 5 to every value results in a SD no diff from the original

3) when you multiply or divide by a constant, the result is as follows:

~measures of both central tendency and variability change; multiplying each value by 5 results in a mean and SD 5 times larger than the originals

4) when you add, subtract, multiply, or divide by a constant, the results are as follows:

~the shape of the distribution remains the same; adding 5 to every value in a skewed distribution results in a similarly skewed dist

~correlations w/ other variables remain the same; adding 5 to every value in one variable does not alter its relationship w/ another variable

Front 22

Standard Score (Z-score)

Back 22

1) obtained by transforming raw scores to obtain a distribution w/ a mean of 0 and a SD of 1

2) Z-scores indicate how many SDs from the mean scores lie (e.g. a z-score of -1.5 indicates a raw score 1.5 SDs below the mean)

3) permits comparisons across diff measures & tests (e.g. students in diff classes performed equally well relative to other students in an exam if both have the same z-score)

4) transforming raw to z-scores does not alter the shape of the new distribution from that of the original dist (i.e. the dist of z-scores will be identical to that of the raw scores)

5) distribution falling on either side of the mean w/in one, two, and three SDs from the mean

~50% of the distribution falls on either side of the mean

~34% btw the mean and one SD

~14% btw one and two SDs from the mean

~2% btw two and three SDs from the mean

6) Other "standard scores"

~T-scores (raw scores standardized to a distribution w/ a mean of 50 and a SD of 10)

~ETS scores (raw scores standardized to a distribution with a mean of 500 and SD of 100)

~why are Standard and Scaled scores not listed??

Front 23

Percentile Rank - Percentage score, percentile rank

Back 23

Percentage Score - number of items out of a total # (e.g. 93 out of 100 = 93%)

Percentile Rank

1) Percentage of scores in a distribution that fall below a particular raw score (e.g. 84% of a distribution is below the 84th percentile)

2) Uniform distribution: equal # of values are to be expected for any given percentile rank

3) changes in scores in the middle of a distribution (where most values are "clumped") are associated w/ larger changes in percentile rank than at extreme ends

~percentile rank increases from the 50th to the 84th percentile when going from the mean to one standard deviation above the mean

~percentile rank increases from the 84th to the 98th percentile when going from one to two standard deviations above the mean

4) Can determine percentile rank from standard score and vice versa

~84% of a distribution falls below a z score of 1 or a t score of 60 (i.e. one SD above the mean); as such, either of those two standard scores may also be referred to as the 84th percentile

~The 84th percentile falls one SD above the mean; as such, one may simply convert to a standard score (e.g. z-score = 1, t score = 60)

Front 24

Frequency Distribution

Back 24

1) Graph presenting variable values on the x (horizontal) axis and frequency of those values on the y (vertical) axis; takes various shapes

2) Normal distribution

~bell shaped

~unimodal (one peak)

~symmetrical (mirror images on either side of mean)

~mean = median = mode

~68% of values w/in one SD of mean, 95% w/in 2 SDs, 99.7% w/in 3 SDs

3) Skewed

~values bunched on one end, tapering off to the other side

~positively skewed if trailing off to the right (positive) side; negatively skewed if trailing off to the left (negative) side

~mean pulled away from mode in direction of skew

~avoid mean as measure of central tendency for skewed distributions

4) Bimodal: two modes or peaks, one on either side of the distribution

5) Uniform: equal frequencies across the distribution (i.e. a block)

6) J-shaped: skewed, but w/out a tail on the side of the distribution w/ the mode

Front 25

ANOVA (Analysis of Variance) - def, One-Way

Back 25

Definition - Measures the variance in populations or samples

One-Way ANOVA

1) Tests for differences in one DV across multiple leves (i.e. conditions) of one IV

2) Omnibus null hypothesis: All of the means are equal (IV has no effect on DV)

3) If rejecting null, we conclude that not all of the means are equal (IV has an effect on DV)

4) does not state which means differ (must conduct post-hoc tests)

Assumptions

1) Independence of observations

2) Homogeneity of Variance - variance within each of the populations is equal.

3) Normality

F-Ratio

1) Variance btw groups (error plus tx) divided by variance w/in groups (error)

2) MSB / MSW (Mean Squared Between / Mean Squared Within)

3) Ratio of approx 1 signifies a lack of tx effect (i.e. w/ no tx effect, no tx variance; with no tx variance, the F-Ratio is formed by error variance divided by error variance, yielding a value of 1)

4) if ratio is sig larger than 1, may conclude means are farther apart than what would be expected from sampling error alone, and thus, IV affects DV

5) If not, null hypothesis retained, conclude that the variability btw sample means can be accounted for by sampling error, that is, IV does not affect DV

Post-Hoc Tests

1) Sig ANOVA indicates mean diffs exist; post-hoc tests tell specifically where the diffs occur

2) examination of which means diff from which

3) Examples:

~Scheffe's test: conservative; provides more protection against Type I errors and incr Type II likelihood

~Tukey's HSD: best for protection against Type 1 errors

~Fisher's LSD: Liberal

Front 26

ANOVA (Analysis of Variance) - Factorial (or n-way)

Back 26

1) n represents number of IVs, or factors

2) used when examining effects of two or more IVs

3) Example: 2x2 design w/ two IVs, each w/ 2 levels (e.g. sex: female/male; instruction method: novel/traditional)

~main effects: diff across sexes, diff across inst method

~interaction: diff effect of one IV at diff levels of the other (e.g. women achieve higher scores than men, but only w/ the novel method)

~w/ an interaction, interpret main effects w/ caution (i.e. story more nuanced than simply "women score higher than men")

~when graphing means using separate line for diff levels of one IV, parallel lines indicate the lack of an interaction effect

Front 27

ANOVA (Analysis of Variance) - mixed design

Back 27

1) multiple IVs incl both w/in subjects (e.g. time) and btw subjects (e.g. condition) factors

2) Example: pre- and post-tests w/ control cond

Front 28

ANOVA (Analysis of Variance) - ANCOVA

Back 28

1) Covariate (extraneous variable) is used to account for a portion of the variance in the DV

2) continuous

3) must be measured prior to IV to ensure independence of tx

4) must be correlated w/ the DV

5) ultimate goal: reduce error variation

Front 29

ANOVA (Analysis of Variance) - MANOVA

Back 29

1) Includes more than one DV

2) if sig effect is obtained, usually follow w/ univariate ANOVAs for each DV in order to interpret

3) Advantages: protects against inflated alpha from numerous ANOVAs; w/ multiple DVs, might reveal effects not found by sep ANOVAs

4) Disadvantages: more complicated design; ambiguous as to which IV affects which DV; increases in power perhaps offset by loss in degrees of freedom

Front 30

ANOVA (Analysis of Variance) - MANCOVA, Discriminate Function Analysis

Back 30

MANCOVA - As w/ an ANCOVA, one or more covariates are added in order to reduce error variation

DFA - used to classify inds into groups based on variables such as age, sex, and level of education

Front 31

Variance

Back 31

1) A measure of variability

2) the square of the SD (V = SD squared or SD = sq root of V)

3) larger for wider distributions, smaller w/ tighter spreads

4) w/ a sample size N, (N-1) in the denominator of the sample variance in order to correct for bias

Front 32

Pooled variance (t-tests)

Back 32

1) Weighted average of two sample variances

2) each sample variance multiplied by degrees of freedom (sample size minus one), then added to create weighted variance

3) weighted variance divided by degrees of freedom for independent samples t-test (total sample size minus two) to obtain pooled variance

4) assumes equal population variances

5) provides better estimate of population variance than either sample alone

6) each sample variance may differ from its corresponding population variance

7) by assuming equal population variances, both sample variances estimate the population variance, and thus a weighted average of both obtains a pooled estimate based on more info than either sample along could produce

https://www.youtube.com/watch?v=XG9ZFH8wwog

Front 33

Moderating and Mediating Variables

Back 33

Moderating

1) a variable that affects the magnitude or direction of the relationship btw the IV and the DV

2) w/ correlations, a moderator is a third variable that affects the correlation btw the IV and the DV (e.g. a therapy has greater effect on depression as age increases)

3) w/ ANOVAs, the interaction btw a moderator and the IV affects the DV (e.g. a therapy has an effect on depression, but only among men)

Mediating Variable

1) a variable explaining the process by which the IV affects the DV (e.g. a therapy affects depression by creating a more positive self-image, which then reduces depression) - MORE INDIRECT

Front 34

Chi-Square

Back 34

1) Examines frequency distribution of categorical variables, such as political party affiliation or eye color

2) non-parametric test that does not require normality

3) goodness of fit: one-way Chi-Square test for examining frequency distribution of one independent variable

~may use expected frequencies given by even split of sample size N across categories (e.g. w/ four categories, expect five observations in each, given 20 observations)

~may also use expected frequencies derived from knowledge of a comparison population (e.g. working on the assumption of 50% of ppl have brown eyes, 30% have blue eyes, and 20% have green or other, one would expect a sample of 50 ppl to produce 25 ppl w/ brown eyes, 15 w/ blue eyes, and 10 w/ green or other-colored eyes)

4) Test for independence: Two-way Chi-Square test for examining contingency table for two variables to determine whether they are independent (i.e. unrelated)

~if a rel does exist, the freq dist of one variable will depend upon the other (e.g. more ill ppl among those exposed to a certain risk factor)

~expected values determined by multiplying column total (e.g. # of ill ppl) by row total (e.g # of ppl exposed to risk factor) and then dividing by the total # of ppl observed (N)

~assumes independent observations; each person appears once and only once in the table (e.g. a blue-eyed person would be counted once in the category "blue eyes", whereas an ill person who had been exposed to a risk factor would be counted only in the cell formed by the intersection of the ill and exposed categories)

~requires cell counts, not percentages

~requires expected cell counts of at least five

5) df: estimate of the # of categories used to organize data

Front 35

Agonist, Antagonist

Back 35

Diff neurotransmitters are implicated in various psychological disorders

1) some disorders are caused by excessive levels of certain neurotransmitters, while others may result from deficits

2) psychotropic meds exert their effects at the synaptic cleft btw neurons (the point of neurotransmitter release by one cell and receipt by the adjacent cell)

Agonists - medication that enhance synaptic transmission and increase post-synaptic effects

Antagonists - meds that inhibit synaptic transmission and decrease post-synaptic effects

Competitive agonists and antagonists

1) molecules that share structural similarities w/ a particular neurotransmitter

2) mimic the neurotransmitter and bind to the same site on receptors to which the neurotransmitter typically binds

3) Comp agonists - incr post-synaptic effects by mimicking the neurotransmitter's effects

4) Com antagonists - occupy the binding site and prevent the neurotransmitter from binding, which results in reduced transmission and activity

Noncompetitive agonists and antagonists

1) typically do not share structural similarities w/ a neurotransmitter b/c they bind to sites on the receptor that are different from those to which the NT binds

2) when bound to a receptor site, they alter its shape

3) this changes the receptor's affinity for the NT

4) affinity can be desc'd as the force of attraction btw the NT and the receptor, and refers to the potency w/ which a NT binds to the receptor site; greater affinity results in more potent binding and vice versa

5) noncompetitive agonists incr the receptor's affinity for the NT, resulting in incr binding and greater post-synaptic effects

6) noncompetitive antagonists decr the receptor's affinity for the NT, causing decreased binding and PS effects

Direct agonists and antagonists - meds that exert their effects by mimicking the NT and binding to its receptors

Indirect - meds that exert their effects by indirectly altering NT release, uptake, or binding

Inverse agonists - have the opp pharmacological effect to that produced by a NT

Partial agonists - have a lesser effect than those produced by a NT

Medical appl of agonists and antagonists

1) older gen drugs (e.g. MAOIs) have lower specificity, and are thus assoc w/ more side effects

~highly specific drugs (e.g. SSRIs) are designed to minimize these probs (i.e. fewer side effects)

2) drugs can be designed to be specific to a particular receptor (e.g. target D2, but not other dopamine receptors)

Front 36

Drug Classification

Back 36

1) Often utilized in psych tx to control symptoms & allow psychotherapy

2) various classification systems incl chemical structure, clinical effect, and pharmacologic mechanism

3) clinical effect classification useful to clinicians: groups the drugs under the effects they appear to exert on the symptoms of certain disorders

Front 37

Drug Classification - Antidepressants

Back 37

1) SSRIs (selective serotonin uptake inhibitors): prevent the reuptake of serotonin from the synapses, prolonging its avail and effects in the brain

~popular SSRIs: Prozac (fluoxetine), Paxil (paroxetine), Zoloft (sertraline), Celexa (citalopram), Lexapro (escitralopram), and Luvox (fluvoxamine)

~prescribed for depression, anxiety, OCD, eating disorders, and also aggressive beh, irritable bowel syndrome, and fibromyalgia, erectile sexual disorders

~ends in INE or PRAM

2) SNRIs (serotonin-norepinephrine reuptake inhibitors): incr both serotonergic and noradrenergic activity

~Effexor (venlafaxine) and Cymbalta (duloxetine)

~ends in INE

3) TCAs (tricyclic antidepressants): exert a general inhibitory effect on the reuptake of all monoamine NTs, incl norepinephrine, dopamine, and serotonin

~many side effects and dietary restrictions; typically prescribed only if SSRIs are ineffective

~Tofranil (imipramine) and Anafranil (clomipramine)

~TCAs treat depression, anxiety, OCD, and neuropathic pain

~ends in INE

4) MAOIs (monoamine oxidase inhibitors): inhibit the activity of monoamine oxidase, which breaks down monoamines and thereby increases serontonin, dopamine, and norepinephrine levels in the brain. They are particularly effective in the tx of atypical depression and have been found to help ppl quit smoking. In general, however, b/c they cause more side effects and have more drug and dietary interactions than SSRIs and TCAs, they are typically reserved as last-option antidepressants

~MAOI brand names include Nardil (phenelzine), Parnate (tranylcypromine), and Emsam (selegiline, the newly dev patch)

Anti-obsessional

1) this is a sub-category of the main antidepressant list, emphasized b/c of their effects on obsessive thoughts (via serotonergic action)

2) It is mostly comprised of SSRIs and includes Prozac, Zoloft, Paxil, Luvox, Celexa, Lexapro, and also the TCA Anafranil (clomipramine)

Front 38

Drug Classification - Mood Stabilizers and Anti-Convulsants

Back 38

1) This is a class of drugs commonly used to treat bi-polar disorder

2) the anti-convulsants were traditionally manufactured for the tx of epileptic seizures, but subsequently found useful as mood stabilizers

3) examples include Eskalith and Lithonate (lithium carbonate), Symbiax (a synthesis of olanzapine and fluoxetine), Tegretol (carbamazepine), Trileptal (oxcarbazepine), Depakote (divalproex), Lamictal (lamotrigine), Topamax (topiramate) and Gabirtril (tiagabine)

Front 39

Drug Classification - Psychostimulants

Back 39

1) Widely used in the tx of ADHD

2) includes "blockbusters" Ritalin and Concerta (methylphenidate), Dexedrine (dextroamphetamine) and Adderall (d- and I-amphetamine)

Front 40

Drug Classification - antipsychotics

Back 40

1) These meds are mainly used to tx psychotic illness; they can also be used as adjuncts for depressive disorders when appropriate

2) this class can be further divided in terms of low and high potency, as well as typical (traditional) and atypical (newer)

~low potency only means that a higher dosage is needed for the desired effect, and does not have to do with side effects and sedative power

~low potency antipsychotics are known to have significant side effects and sedative power

~low potency meds include Thorazine (chlorpromazine), Mellaril (thioridazine), Clozaril (clozapine) and Seroquel (quietapine)

~high potency meds include Loxitane (loxapine), Stelazine (trifluoperazine), Prolixin (fluphenazine), Navane (thiothixene), Haldol (haloperidol), Risperdal (risperidone), Zyprexa (olanzapine) and Abilify (aripiprazole), among others

Front 41

Drug Classification - Anti-Anxiety

Back 41

1) can be further divided into Benzodiazepines and other anti-anxiety agents

2) they are widely used to treat both anxiety and mood disorders, and can be adjuncts in antipsychotic regimes

3) benzos include Valium (diazepam), Librium (chlordiazepoxide), Klonopin (clonazepam), and Ativan (lorazepan).

4) other anti-anxiety agents include BuSpar (buspirone), Neurontin (gabapentin), and the beta-blocker Inderol (propanolol)

Front 42

Drug Classification - Hypnotics

Back 42

This class includes mostly "sleepers" such as Dalmane (flurazepam), Doral (quazepam), Restoril (temazepan), and the newer Ambien (zolpidem), Sonata (zaleplan) and Lunesta (eszoplclone), as well as the antihistaminic Benadryl (diphenhydramine) for its sedative side effects

Front 43

Drug Classification - over the counter and natural / herbal meds

Back 43

1) These products have been shown to tx psych conditions in clinical trials

2) they include St. John's wort (an herbal product shown to tx depression and anxiety), SAM-e (depression) and Omega-3 (depression and bipolar disorder)

3) use judgment when using them -why?

Front 44

Antidepressants - SSRIs

Back 44

SSRIs

1) Serotonin agonists: prevent the reuptake of SE from synapses, thereby prolonging the effects of SE

2) Indication:

~most popular tx for serontonergic imbalances

~primarily used to tx depression, some anxiety disorders, OCD, aggressive beh, irritable bowel syndrom, fibromyalgia, eating disorders (e.g. Bulimia is assoc w/ low levels of SE)

3) Examples of SSRIs: Prozac (fluoxetine) is one of the most common SSRIs used to tx dep, Paxil (paroxetine), Zoloft (sertraline), Celexa (citalopram), Lexapro (escitralopram), and Luvox (fluvoxamine)

4) SSRIs produce fewer side effects than MAOIs and TCAs, making them a more popular option

5) side effects: dry mouth, vivid dreams, constipation, sexual dysfunction, nausea, drowsiness, dizziness, changes in appetite, weight loss/gain, suicidality, and liver or kidney impairment

~side effects tympically diminish w/in the first few weeks of taking the meds

~SSRI regimens are usually started at a low dose and gradually incr to reduce the risk of side effects

~abrupt early discontinuation of the meds is not advisable

MOST contain either "ox" or "pram" in the generic name

Front 45

Antidepressants - SNRIs

Back 45

SNRIs

1) work to incr both SE and Norepinephrine (Noradrenaline) activity

2) SNRIs are used to tx anxiety disorders, ADHD, and neuropathic pain

3) SNRIs include Effexor (venlafaxine), Cymbalta (duloxetine), and Wellbutrin (bupropion)

4) side effects similar to SSRIs

MOST contain "ax", "ox" in generic name

Front 46

Antidepressants -TCAs

Back 46

TCAs

1) class of antidepressant meds used since the 1950s (1st gen)

2) exert general effect on the reuptake of all monoamine NTs, incl norepinephrine, DA, & SE

3) have affinity for muscarinic receptors - parasympathetic effects (such as a slowed heart rate and increased activity of smooth muscle) - Muscarinic acetylcholine receptors, or mAChRs, are acetylcholine receptors that form G protein-coupled receptor complexes in the cell membranes of certain neurons and other cells. Muscarinic receptors are so named because they are more sensitive to muscarine than to nicotine.

4) produce fewer side effects than MAOIs

5) not as effective as SSRIs (?) CONTROVERSIAL, depends on how "efficacy" is defined

6) used in the tx of depression, anxiety, OCD, and neuropathic pain

7) commonly used TCAs incl imipramine (Tofranil) and clomipramine (Anafranil) - both name brands end in "nil"

8) side effects:

~generally more severe than SSRIs

~incl dry mouth, blurred vision, constipation, diff w/ urination, and hyperthermia

~other side effects incl anxiety, drowsiness (somnolence), confusion, incr appetite, decr sexual ability, and some cardiovascular effects

~similar to SSRIs, side effect typically disappear in a few weeks and are better tolerated when the med is slowly incr

~TCA toxicity can lead to severe cardiovascular and neurologic effects and is often fatal

Front 47

Antidepressants - MAOIs

Back 47

MAOIs

1) first class of antidepressants to be dev

2) MAOIs inhibit the activity of monoamine oxidase, which breaks down monoamines and thereby incr SE, DA, and NE levels in the brain

3) cause more side effects and have more drug and dietary interactions than SSRIs and TCAs; they are typically reserved as a last option antidepressant

~the recently dev MAOI skin patch (transdermal patch) reduces these risks - Selegiline (Emsam)

4) interactions btw MAOI chemicals (e.g. tyramine) in certain foods & other drugs may result in hypertensive crisis; thus, ind's taking MAOIs should avoid consuming foods that contain high levels of tyramine, such as aged cheese, wines, liver, beer, and soy products

5) MAOIs are particulary effective in the tx of atypical depression and have been found to help ppl quit smoking

6) some common MAOIs incl isocarboxazid (Marplan), iproniazad (Irprozid), and phenelzine (Nardil)

SOME generic names inclu "zid" or "zad"

Front 48

Antidepressants - SE syndrome

Back 48

1) both SNRIs and MAOIs can cause SE syndrome if combined with SSRIs, typtophan, illicit substances, or other over the counter meds like St. John's Wort (itself thought to be an MAOI), which are known to incr SE levels

2) SE syndrome is a severe and potentially fatal condition that incls agitation, restlessness, rapid heart rate, dilated pupils, loss of muscle coordination, and cognitive symptoms (e.g. hallucinations, confusion)

Front 49

Antidepressants - use of antidepressants in other populations

Back 49

1) both panic disorder and chronic pain probs are commonly tx prophylactically w/ antideps

2) there is little concern in prescribing antideps to someone who has schizoprenia and is depressed

3) pregnant women: neonates born to mothers who were taking SSRIs during pregnancy may suffer from SSRI w/draw and may be at a greater risk for certain perinatal conditions

4) studies have shown that Prozac, Zoloft, and Paxil may cause incr agitation, aggression, psychosis, and suicidality in some inds, particularly adolescents

5) although the risk of cardiovascular accidents is lower in SSRIs, there still remains some risk; therefore, use of SSRIs for inds w/ coronary artery disease is contraindicated

*6) most sig concern in prescribing antidepressants is for someone who has bipolar disorder, as admin of an antidep may trigger a manic episode

Front 50

Antipsychotic Medications and Side Effects - Psychosis

Back 50

Psychosis

1) A sudden or gradual loss of contact w/ reality

2) Characterized by the development of delusions, hallucinations, and disorganized behavior

3) Psychosis can be affiliated w/ a wide range of disorders incl bipolar disorder, major depressive disorder, brain tumors, seizure disorder, schizophrenia, dementia, and stroke

Front 51

Antipsychotic Medications and Side Effects - Mechanism of Action, Tx of psychotic symptoms w/ "Typical" meds

Back 51

Mech of Action - block D2 dompamine receptors

Tx w/ "Typical" Antipsychotic Meds

1) focused on symptom mgmt

2) antipsychotic med side effects can be severe:

~patients discontinue meds due to side effects

~akathisia is the most common; involves feeling dsyphoric, restless, tapping of feet, rocking, and shifting

Early "typical" antipsychotic meds

1) neuroleptic meds such as Thorazine and Haloperidol

2) effective in controlling positive symptoms, incl hallucinations and delusions

3) often ineffective in tx negative symptoms, incl anhedonia and dulled emotionality

4) typical antipsychotics are thought to block DA (and other) receptors, leading to a range of unintended side effects:

~anticholinergic side effects include dry mouth, diff urinating, constipation, blurry vision

~adrenergic side effects incl postural hypotension and sexual dysfunction

~histaminergic side effects incl incr sedation and gradual weight gain

~extrapyramidal side effects (EPS): due to interference w/ the DA system, parkinson-like symptoms, tremor, body rigidity, akinesia (impaired ability to initiate movement)

*EPS = Acute dystonic reactions: muscular spasms of neck, jaw, back, extremities, eyes, throat, and tongue; highest risk in young men; occurs w/in a few days of tx onset; postural abnormalities, diff swallowing (dysphagia), uncontrollable eye movements; blank stares (oculogyric crisis);

~Akathisia: A feeling of internal motor restlessness that can present as tension, nervousness, or anxiety; refers to a sense of distress and restlessness; rocking back and forth, shuffling, pacing, or other repetitive movements

~Pseudoparkinsonism: drug-induced parkinsonism (rigidity, bradykinesia, tremor, masked facies, shuffling gait, stooped posture, sialorrhoea, and seborrhoea; greater risk in the elderly). Although Parkinson's disease is primarily a disease of the nigrostriatal pathway and not the extrapyramidal system, loss of dopaminergic neurons in the substantia nigra leads to dysregulation of the extrapyramidal system. Since this system regulates posture and skeletal muscle tone, a result is the characteristic bradykinesia of Parkinson's.

~Tardive dyskinesia: involuntary, uncontrollable, restless movements, muscle movements in the lower face and distal extremities (facial grimaces, tongue protrusions, eye blinking, and limb movements); this can be a chronic condition associated with long-term use of antipsychotics.

~Neuroleptic Malignant Syndrome (NMS): potentially fatal syndrome associated w/ the use of neuroleptic or antipsychotic meds; symptoms incl severe muscle rigidity and hyperthermia, diaphoresis, dysphagia, tremor, incontinence, changes in consciousness, mutism, tachycardia, elevated/labile blood pressure, leukocytosis, and muscle injury; onset of NMS usually occurs w/in 4 weeks of beginning a neuroleptic med

Front 52

Antipsychotic Medications and Side Effects - Tx w/ "Atypical" antipsychotics

Back 52

1) Atypical anti-psychotics serve as SE and DA antagonists

2) incl clozapine, risperidone, quetiapine, olanzapine, ziprasidone

3) tx negative symptoms

4) incr effectiveness in tx-resistant patients

5) produce fewer EPS effects

6) Clozapine is associated w/ a risk of agranulocytosis and requires carefule monitoring of blood levels (registry); agranulocytosis has not been associated w/ other atypical antipsychotics

Front 53

Cluster sampling

Back 53

1) sampling technique involving naturally occurring groups (clusters)

2) population is divided into clusters and some clusters are randomly selected for inclusion in the sample

3) study information collected from all elements w/in clusters is included in the sample

4) clusters should be internally heterogeneous yet relatively homogeneous btw clusters (i.e. variation should be more w/in rather that btw clusters)

5) diff btw cluster and stratified sampling

~stratified sampling: sample drawn from each stratum; main obj is improved precision

~cluster sampling: only elements of randomly selected clusters are studied: main objs are to imp sampling efficiency and reduce cost

~multistage sampling: more complex form of cluster sampling

~population divided into strata at highest level, then sample is drawn and stratified at lower level, procedure repeated until arriving at lowest hierarchical level

Front 54

Random Assignment & Random Selection

Back 54

Random Selection

1) drawing a sample from a population in such a way that each member has an equal probability of being selected

2) supports external validity

~findings from a sample representative of a pop are more generalizable

Random Asssignment

1) assigning participants to an experimental condition in such a way that they all have a equal chance of appearing in any given condition (e.g. flip a coin to determine assignment to the control or experimental groups)

2) supports internal validity

~renders conditions equivalent (i.e. similar), which means that the IV should be the only factor varying among conditions

3) If random assignment is not poss due to pre-existing groups (e.g. classrooms or schools), a quasi-experiment is in order

Front 55

T-test

Back 55

1) determines whether one mean equals a hypothesized value or whether two sample means are equal

2) the t statistic is calculated by dividing the diff btw one sample mean and a hypoth value or the diff btw two sample means by the Standard Error of that particular diff statistic

3) if the t statistic is large enough, one may declare the diff to be sig (the mean is diff from the hypoth value, or the two means are diff)

4) the t-test is more powerful (i.e. more likely to reject the null hypoth) with:

~larger sample sizes

~larger mean diff

~smaller sample variation

5) specific t-tests

~one sample t-test: tests the hypoth that a single sample mean is diff than a specific hypoth value

~Independent samples t-test: tests the hypoth that two unrelated samples are diff from each other

~related or dependent samples t-test: tests the hypoth that the diff btw two related samples (e.g. pre and post-test scores, scores of siblings) is not equal to zero (i.e. samples have diff means)

Front 56

Sample size - Sample, larger preferred

Back 56

Sample

1) a subset of a pop intended to provide a small snapshot representing the whole pop

2) random sampling (i.e. every member of a pop has an equal prob of being selected) affords the best opp of obtaining a repr sample (i.e. a sample matching the characteristics of the pop)

3) sample size: # of observations (e.g. ppl) in a sample, denoted by N

Larger sample sizes preferred

1) to ensure that the sample adequately represents the pop

~example: if sampling two ppl from a neighborhood w/ an equal mix of women and men, one is just as likely to find two women or two men as one is to find one of each

~if sampling 25 ppl from a neighborhood, one is far less likely to obtain a sample made up entirely of either women or men

2) to reduce sampling error - a more repr sample provides statistics closer to the corresponding pop parameter

3) to incr statistical power - with fewer errors (i.e. noise), diffs or relationships are more likely to be deemed stat sig

Front 57

Central Limit Theorem

Back 57

Increasing the size of the random sample N drawn from a pop will cause the distribution of the sample means to:

~form an incr normal dist w/ a mean equal to the pop mean

~also w/ a SD (i.e. the Standard Error of the mean) equal to the sample SD divided by the square root of N

Front 58

Correlation

Back 58

Correlational (observational) study

1) examines rel btw un-manipulated variables

2) measures association; does not estab cause & effect (x may cause y, y may cause x, or a third variable z drives both)

3) pearson's r: measure of linear rel btw two variables; ranges from -1.0 (perfect negative rel) through 0 (no rel) to +1.0 (perfect positive rel)

4) assumptions: independent observations, linear rel, bivariate normality (joint distribution of both variables together are normal)

5) graphically presented w/ a scatterplot

6) weakened by restriction of range

7) susceptible to bivariate outliers (observations far from the mean of both variables)

8) suppressor variable may hide correlation

9) partial correlation: correlation btw x and y after removing variability in each shared w/ a third variable z

10) semi-partial correlation: correlation btw x and y after removing variability in x (and only x) shared w/ a third variable z

11) coefficient of determination: r-squared

~proportion of variability in one variable shared w/ another

12) other correlation coefficients

~point-biserial: one continuous variable, one dichotomous variable (e.g. female/male); alternative or supplement to t-test

~Biserial: two continuous variables, one made into an artificial dichotomous variable

~phi: two dichotomous variables

~tetrachoric: two dichotomized variables, used in Item Response Theory

~contingency: two nominal variables

~Spearman's rho: two ordinal variables

~Eta: nonlinear rels btw two variables

~Canonical: two sets of variables, one repr multiple IVs, the other multiple DVs, examines many-to-many rather than one-to-one or many-to-one rels; produces multiple correlation coefficients (the first accounting for the largest portion of the relationship)

13) logistic regression is used to predict discrete outcomes

~

Front 59

Autocorrelation

Back 59

1) rel btw two values of the same variable measured at diff times

2) correlation btw measurements of a dependent variable taken at diff time from the same subjects

3) in regression analysis (analysis of the rel btw the IV and the DV), it tends to underestimate error terms, inflating t-values and reducing p-values (i.e. so the results are more likely to be deemed stat sig)

4) if the researcher can determine the prediction error in one observation, then a good guess about the error in a linked observation can be made

5) determining those links also determines autocorrelation, which then allows more info to be extracted from the data

6) in a time-series design, autocorrelation serves two purposes: a) to detect non-randomness in the data (actual sig); b) to identify an appropriate time-series model if the data are not random

7) consistent rels in time-series data can be used to predict future values in the series - some ability to forecast itself

Front 60

Effect Size

Back 60

1) measures determine practical rather than statistical sig ("is the effect large enough to matter?" vs "does the effect exist?")

2) used in meta-analyses to combine findings from multiple research studies due to its independence from sample sizes

3) specific effect size measures vary by context

4) correlation (r-squared): proportion of variation in one variable accounted for by the linear relationship w/ another

5) chi square (Cramer's phi): strength of rel btw two variables in a contingency table

6) t-test: (cohen's d) - diff btw two group means in terms of SD (control group or pooled)

7) ANOVA (eta squared or omega-squared): proportion of variation in DV accounted for by the IV

Front 61

Statistical significance

Back 61

1) results of an analysis reflecting more than a chance deviation, allowing for the conclusion that an effect (relationship, mean diff) likely exists

2) p-value is the conditional probability of obtaining the obtained results assuming that the null hypoth is true

3) p-value compared to a conventionally determined alpha value to determine stat sig

~alpha is the max acceptable prob of making a Type I error (erroneously declaring an effect)

~if p > alpha, it is too likely that the results were due to chance occurrence, and thus one cannot conclude that there is an effect

~if p < alpha, it is relatively unlikely that the results occurred due to chance, and thus the existence of an effect is concluded

4) Exp: does our test prep course affect math SAT scores?

~alternative, or research, hypoth is that there is an effect; the mean of those who took the prep course is not equal to 500

~Null hypoth is that there is not an effect, the mean of those who took the prep course is the same as those who did not (500)

~assume the null hypoth for a moment: if the null were true, what is the prob (p-value) of having obtained results at least as extreme as those that we obtained?

~if p-value (e.g. .34) is greater than alpha (e.g. .05), the obtained results were not sufficiently diff to conclude a diff attributable to anything more than chance variation

~if p-value (e.g. .034) is less than or equal to alpha (e.g. .05), the obtained results were sufficiently diff to conclude a diff beyond chance variation (i.e. a true effect)

5) Type I error: conclude an effect exists when the obtained results ere actually due to chance

~the prob of committing a Type I error is alpha

6) Type II error: conclude that no effect exists when the results were actually due to chance

~the prob of committing Type II error is beta

7) Power, or the prob of concluding an effect exists if there really is an effect (1 - beta), can be incr by:

~incr the magnitude of effect

~decr error variance

~incr sample size

~incr alpha (i.e. run a less conservative test)

~power analysis: calculate the sample size required to capture an effect

~conduct a one-tailed, or directional, test (i.e. focus specifically on higher or lower end of the null distribution) instead of a two-tailed, or non-directional test, which broadly looks for an effect

8) Familywise alpha

~running multiple tests, each with its own alpha, results in a familywise alpha for the entire set approx equal to the product of all the alphas

~exp: running 5 t-tests, each w/ an alpha of .05, yields a familywise alpha of .05 x 5, or .25

~this can be corrected (e.g. Bonferroni adj), or an alternative test can be conducted (e.g. MANOVA to replace multiple ANOVAs)

Front 62

Confidence Interval

Back 62

1) range of values centered at sample statistic used to estimate the pop parameter w/ a confidence of (1-alpha) percent

2) if 100 confidence intervals for a pop mean were created from 100 samples, 95 would be expected to contain the population mean when using an alpha value of .05

3) Exp: to create 95% CI for a sample mean

~center at sample mean

~add/subtract from that point estimate the critical values of the test statistic multiplied by the standard error of the sample statistic

~sample mean = 20, standard error = .10, using critical z-values (rounded for simplicity) given a large sample size and an alpha of .05

~20 + (2 x .10) = 20.2

~20 - (2 x .10) = 19.8

~95% CI ranges from 19.8 - 20.2

4) the sample has produced an est of the pop mean that may be used instead of, or in add to, hypoth testing (i.e. determining the prob that the pop mean is a part value)

Front 63

Coefficient of (Multiple) Determination (R-squared) -

Back 63

1) expresses the proportion of Variation in a DV accounted for by the IV

2) reflects the reduction in error achieved by using one or more IVs in predicting the DV as opposed to only using the DV's mean in making estimates

3) not good for sample to sample comparison b/c of divergent variance in the DV w/ each sample (i.e. differing values of total variation)

4) higher w/ more variables

~corrected w/ adjusted r-squared: more conservative and lower estimates of the variation in the DV are accounted for

5) No matter how high the r-squared value, correlations are a matter of association and not causation

R-squared = explained variation (SSR) / Total variation (SST) - how closely points (data) are to the regression line

R-squared = ALSO equals square of correlation coefficient

Front 64

Proband

Back 64

1) AKA "patient zero" or the "index case"

2) the first family member to seek prof attention for a disorder

3) the family research method focuses on patterns of a disorder among Probands and relatives

4) probands share a known # of genes w/ family members

~if a genetic predisposition to a disorder is found, the likelihood that relatives of Probands will receive the same diagnosis is higher than that observed in the gen pop (i.e. concordance btw shared genes and a disorder)

~example: the rate of diagnosis of schizophrenia is approx 1% in the gen pop; whereas 10% of first degree relatives of Probands receive the diagnosis

Front 65

SEM formula / How to calculate CI

Back 65

SEM = pop SD / sq root of sample size

sample = 100

SD = 10

mean = 50

SEM = 10 / sq root of 100 = 10/10 = 1

CI (95%) = 50 +/- (2 x 1) = 50 +/- 2 = 48 - 52

Front 66

Actuarial Data - Def, Grove and Meehl

Back 66

Def - Predictions based on statistical info (objective) rather than judgment (subjection)

Grove and Meehl

1) compared actuarial data to clinical judgement

2) related to calculation of birth and death risks

3) predictions form actuarial data are equal to or better than predictions from clinical judgments; clinical judgments rarely outperformed actuarial data

4) meta-analysis of 136 studies: 8 favored clinical judgment, 64 favored actuarial, 64 equivalent

5) later meta-analyses: effect sizes for actuarial data 10% better than clinical judgment; actuarial predictions of delinquent and criminal behs more accurate than clinical judgments

6) ruled out alternative explanations such as examiner's field of training, length of exp, and task-related exp

Front 67

Protocol Analysis

Back 67

1) Qualitative data analysis method involving verbalization of thoughts that arise while completing a given task

2) assumes that verbally expressing one's thoughts does not alter the sequence of thoughts required to perform a task

3) obtained reports (protocols) are analyzed to gain an understanding of how participants solve probs

4) other poss indicators in addition to verbal reports:

~reaction times

~error rates

~brain activation patterns

~eye fixation sequences

5) strong correspondence btw thoughts and info/objs examined

6) one of the principal research methods in cognitive psychology, cog science, and beh analysis

Front 68

Data reduction techniques

Back 68

1) methods by which the interrelationships btw a set of variables are analyzed to produce a smaller # of dimensions (or factors)

2) PCA (Principal component analysis): analyzes all the variability in a set of observed variables to produce a smaller # of components (or factors)

~factor loading is the correlation between an observed variable and a given factor

~eigenvalue is the amount of variability in the observed variables accounted for by a given value

~eigenvalue is the sum of the squared factor loadings for a given factor

~preferred over PAF (Principal axis factoring) for data reduction. PAF is similar to PCA, but removes variability unique to the ind observed variables and instead analyzes only common variability in producing factors; PAF is preferred over PCA for finding underlying structure

~PCA is a linear combination of variables; Factor Analysis is a measurement model of a latent variable.

~A Factor Analysis approaches data reduction in a fundamentally different way. It is a model of the measurement of a latent variable. This latent variable cannot be directly measured with a single variable (think: intelligence, social anxiety, soil health). Instead, it is seen through the relationships it causes in a set of Y variables.For example, we may not be able to directly measure social anxiety. But we can measure whether social anxiety is high or low with a set of variables like “I am uncomfortable in large groups” and “I get nervous talking with strangers.” People with high social anxiety will give similar high responses to these variables because of their high social anxiety. Likewise, people with low social anxiety will give similar low responses to these variables because of their low social anxiety.

https://www.theanalysisfactor.com/the-fundamental-difference-between-principal-component-analysis-and-factor-analysis/

https://www.sagepub.com/sites/default/files/upm-binaries/19710_784.pdf

3) Cluster analysis: "eyeballs" groups (or clusters) from correlation matrix

~two most strongly correlated variables form nucleus added to cluster

~other variables correlated w/ nucleus added to cluster

~two strongly correlated variables w/ weak correlations for first cluster form nucleus for second cluster

~variables correlated w/ second nucleus added to second cluster (and so on)

4) LTA (latent transition analysis): form of factor analysis for categorical data

~freq used in educational testing and psych measurement

~categorical variables (e.g. responses to a multiple-choic exam such as the SAT) are reduced to latent traits (e.g. academic ability)

~LTA is a longitudinal version of LCA (latent class analysis)

~LTA allows researchers to characterize the memberships to diff groups as well as to predict changes among memberships

https://www.frontiersin.org/articles/10.3389/fpsyg.2018.00675/full

Front 69

Developmental Research

Back 69

1) Assesses changes over a period of time, and 3 designs are typically used

2) Longitudinal: one group is followed for an extended period of time (e.g. first graders tracked for 12 years)

~lengthy time requirement a drawback

~participant mortality due to a # of factors (illness, relocation, etc) a concern

~lack of randomization another pot issue

~history primary a threat to external validity

~can provide valuable qualitative and quant data

3) Cross-sectional: groups at each level are assessed at the same time (e.g. 12 groups of students, one at each grade level, are observed at one time)

~assumes that diffs reflect natural dev (i.e. a longitudinal study would produce similar findings)

~req's much less time than a long. study

~diffs may be due to cohort effects (i.e. group diffs reflect diff exps rather than natural dev)

4) cross-sequential: comb of long. and cross-sectional designs

~diff groups assessed repeatedly over time (e.g. groups of 1st, 4th, 7th, & 10th grade students are assessed over 3 years, thus providing coverage from 1st - 12th grade)

~reduces time required to perform and minimizes assumptions/cohort effects

Front 70

ABAB design

Back 70

1) Single subject design in which a baseline measure of the DV (e.g. depression) is obtained (A) b/f the tx is introduced (B), removed (A), and finally reintroduced (B)

2) If tx has an effect, then the DV will:

~deviate from baseline when it is introduced

~return to baseline when removed

~deviate again from baseline when reintroduced

3) if the DV does not return to baseline after removal of tx, the initial deviation may be attributable to a confounding variable

4) tx is reintroduced after the 2nd baseline in order to:

~further estab the tx effect

~restore benefit of the tx to subject

5) ABA design is common in basic research but does not reintroduce tx after 2nd baseline

6) ABAB design may be altered to incl diff tx: following baseline and tx for therapy 1 (A1B1), study may be cont w/ a subsequent baseline and tx series for therapy 2 (A2B2) and so on

Front 71

Double-blind design

Back 71

1) participants and experimenters are blind (i.e. naive) to exp cond

2) exp's may unconsciously infl participant beh to conform to research hypoth (expectancy effects)

3) subtle cues, such as tone of voice or posture, may bias research results to fit expectations

4) an experimenter ignorant to the exp cond should be unable to infl responses such that they conform to expectations

5) variations:

~single-blind study - part is naive to cond, exp is not

~triple-blind study - in addition to parts and exps, others involved in the research (e.g. pharmacists, statisticians) are naive to assignment

Front 72

Latin Square Design

Back 72

1) counterbalancing technique used for multiple tx designs when order effects are poss and there are too many tx for compl counterbalancing

2) create a reduced # of tx sequences equal to the number of diff tx (e.g. w/ four tx: ABCD, DABC, CDAB and BCDA) from the set of all possible tx sequences

~each tx appears in each serial position only once

3) randomly assign parts to these tx seqs

4) w/ an even # of tx, a balanced Latin Sq design can be used

~each tx appears before and after other tx an equal # of times (e.g. tx A appears twice b/f and twice after tx B)

5) tx and order effects can be analyzed

6) does not completely control for order (not all poss tx sequences used)

Front 73

Single-subject design

Back 73

1) used in beh modification studies in which the intervention may be evaluated w/ a single subject

2) estab causality and minimizes threats to validity

3) large amount of variation in DV represents a potential threat

4) preferred for ethical reasons (conclusions may be drawn w/out removing tx)

5) common applications:

~ABAB design: baseline estab, then tx intro'd, removed, and reintro'd

~multiple baselines: tx effect examined w/ diff behs, ppl, or settings

6) Four key characteristics

~continuous assessment: measured at several diff times b/f and after tx is introduced

~baseline assessment: measurements taken b/f tx in order to estab a pre-existing trend

~stability of performance: measurements made until stable levels are obtained

~diff phases: baseline, tx, and poss addt'l tx phases are used to examine efficacy of an intervention

7) non-statistical evaluation:

~mean changes

~level changes (change btw last measurement of one phase and first measurement of the next)

~slope (or trend) changes

~latency of change (speed w/ which change occurs upon phase change)

8) problems

~autocorrelation: w/ repeated measurements of the same variable w/ the same participant, observations drawn at different times may be correlated

~practice effects

~lengthy time requirements

~potential lack of generalizability (concl drawn from 1 person may not apply to another)