Final Exam prep Flashcards
Why do we have science in psychotherapy? What questions does it address?
Because therapies that are not science-based can be harmful/dangerous
Addresses:
Whether treatment works
Are all treatments equal?
Is the field taken seriously (e.g., medical community refers to us vs. psychics, etc.)
If no support behind methods, other people (e.g., insurance companies, politicians, celebrities) would decide what treatment to use (based on popularity, low cost, etc.)
Less likely to be sued for malpractice
Directs movement towards more effective treatment, or tells us when we can stop looking (i.e., CBT used for panic disorder treatment)
Allows for therapists who don’t practice scientifically to be excluded, or not chosen as therapists
What are some examples of dangerous non-scientific practices?
E.g., “Rebirthing therapy” which resulted in a death
E.g., “repressed memories” - many patients who have now retracted their claims and know that they were false
What was Hans Eysenck’s 1952 study about? What three groups were compared?
“The Effects of Therapy, an Evaluation”
Reviewed 24 outcome studies 1920 - 1950
Compared three groups:
Psychoanalysis
Mixture of therapy
No therapy
What were the results of Hans Eysenck’s 1952 study?
What did he conclude?
Results:
44% of psychoanalysis group improved
64% of mixed therapy group improved
72% of no-therapy group improved within 2 years of symptom onset
Concluded:
The effects of therapy are small to non-existant
What are some criticisms of Hans Eysenck’s 1952 study? (5)
No randomized control trials (RCTs)
Without radomization, comparison groups weren’t equal on a number of variables (outcomes could have been because of a variety of factors)
The no-therapy group wasn’t a good control because they did receive attention, medical care and treatment, etc. (not a true no-therapy group)
Data about efficacy was “authoritatively proclaimed” or based on successful case histories (not based on good, scientific evidence) - often based on case histories
Relied on records across all patients/settings being kept the same way (assumed) but this is probably not true
What were the Smith, Glass and Miller meta-analyses of 1977 and 1980 comparing?
Combined results of 370 (1977) and 475 (1980) studies
Overal effect sizes of therapy overall and specific therapies
What was the mean effect size of therapy overall in the 1980 Smith, Glass and Miller study?
They found that the average patient is better off than __% of the people who are untreated.
What conclusion can we make from these results?
.85
80%
Clear evidence for a significant effect of therapy
What are non-specific factors of therapy?
E.g., an appointment every week, attention, care, having rituals, expectations of improvement, etc.
Dodo Bird Effect
-who coined it
-what does it suggest?
Luborsky (1975) “Everybody has won, and everybody should win a prize” (allusion to Alice in Wonderland quote)
Suggests that all therapies have the same outcome
What did Rachman & Wilson, 1980, say about the Dodo Bird effect?
If the indiscriminate distribution of prizes argument carried true conviction… we end up with the same advice for everyone – “Regardless of the nature of your problem seek any form of psychotherapy.” This is absurd. We doubt whether even the strongest advocates of the Dodo bird argument dispense this advice.
What findings seem to somewhat support the Dodo bird effect?
Some research has shown effect size differences - does not completely support Dodo bird effect (although effect sizes are somewhat similar between some therapies):
Smith & Glass, 1977:
Cognitive 2.38
CBT 1.13
Client-centered .62
Smith, Glass & Miller, 1980
Cognitive 1.31
CBT 1.24
Humanistic .63
What are the criticisms of the Smith & Glass studies which seem to support the Dodo bird effect?
no randomizations, small samples
Shapiro & Shapiro (1982)
- what were they looking at?
- what were the results?
- what does this suggest?
Researchers started looking at how different treatment types affected specific diagnoses which found that there are large effect size differences in this regard:
Anxiety and depression
Behavioural = .74
Cognitive = 1.34
Psychodynamic/humanistic = .40
Phobia
Behavioural = 1.46
Cognitive = .92
Psychodynamic/Humanistic = analysis not conducted
Findings do not support the Dodo bird effect
Reid (1997)
- what were they looking at?
- what were the results?
- what does this suggest?
42 meta-analyses of treatments for depression, insomnia, smoking cessation, and bulimia
75% of the meta-analyses showed evidence of treatment effects
Behavioural treatments (including CBT) superior to non-behavoural treatments
Findings do not support the Dodo bird effect
Wampold et al., 1997
- what were they looking at?
- what were the results?
- what does this suggest?
- what are some criticisms of this study?
Meta-analysis from studies published between 1970-1995 that compared at least two treatments
Found average effect size of only .19 (small difference between treatments)
Support for the Dodo bird effect
Criticisms: most of the studies they analyzed were based on different forms of CBT, not different orientations of treatments (so likely CBT compared with CBT would have a small effect size)
Siev & Chambless, 2007
- what were they looking at?
- what were the results?
- what does this suggest?
All published studies comparing CBT to relaxation for treatment of GAD and panic disorder
GAD: the two treatments were similar
Panic = CBT > relaxation (much better results for CBT)
Support against the Dodo bird effect:
Based on research evidence, what is our conclusion about the Dodo bird effect?
Sometimes the effect sizes can be small but other times (as in the case of panic disorder) the effect sizes can be significant. Summary: sometimes treatments are equal but in some conditions, treatments are not equal
Evidence-based practice emphasizes first using the intervention that has the greatest support, but if the intervention proves to be unsuccessful with a given patient, turning to treatments with less support is entirely appropriate
What could stand in our way of finding out if treatments work and if there are certain treatments that are better than others?
We tend to generalize (e.g., “It probably always works, or works for everyone”)
Might also bias results based on affiliation with the therapy (my school taught me this therapy so it must work)
What if a therapy DIDN’T work? Would we still see change? What are some of the reasons that we might see change for a person, even if the actual treatment didn’t work?
We may tend to think our therapy is working, based on our experience with the therapy
Passage of time, regression towards the mean (possible that simply passage of time affects change or there is regression towards the mean)
What are Empirically Supported Therapies (EST)?
Clearly specified psychological treatment shown to be efficacious in controlled research with a delineated population
APA task force suggests that good treatments must meet a specific criteria and those that meet the criteria would be labelled ESTs
What are the aspects that need to be addressed in research for it to be EST?
Efficacy
Specificity
Effectiveness
Efficiency
Other important considerations:
-Sample Description
-Treatment manual
-Reliable and valid outcome assessment measures
-Appropriate data analyses
What are the rules set out by Chambless & Hollon (1998) for a study to be considered efficacious for an EST?
Controlled research with Random Assignment - called randomized controlled trials (RCTs)
RCT = patients randomly assigned to the treatment of interest OR to one or more comparison conditions
Efficacious:
When a treatment has been found efficacious in at least 2 studies
And… when a treatment has been found efficacious by independent research teams
Possibly efficacious:
Only 1 study
Or… all research conducted by one team
What is specificity? How are studies designed to test specificity?
If the treatment is found to be superior to conditions that control for nonspecific processes, such as receiving attention for being in a treatment group, or expecting that you will change
The treatment must be successful specifically due to the elements of the treatment itself rather than other factors
Using credible sham placebos or other ESTs as a control group is how to test for and determine specificity - then we know that the change/result from treatment is due to the treatment itself and not due to other factors
What are the differences between the efficacy and the specificity of treatments/therapies?
Efficacy:
Determining if the treatment produces change compared to a control
Specificity:
Determining if the treatment itself produces the change or if other variables are producing change
Efficacy:
Has to be replicated (otherwise only possibly efficacious)
The control could be a wait-list, placebo, or another type of treatment (just has to have a control of some kind)
Specificity:
The control has to be a credible sham placebo or another treatment that has already been proven efficacious
Example of efficacy and specificity:
participants - randomly assigned to two levels of independent variable
Experimental condition: write in a journal for 15 min/day for 4 weeks about their emotions that day
Control condition: do nothing - life as usual
Measure: Beck Depression Inventory
What is the efficacy and specificity in this experiment?
Efficacy: random assignment to two conditions and there was change in the experimental condition compared with the control condition
But… was the change because they were writing about emotions or was it simply because they had something to do every day?
Specificity: the control has to be a credible sham placebo or another treatment that already has proven efficacy (e.g., write in a journal for 15/day for 4 weeks about that day or CBT for depression) – if we find that the experimental condition produces a greater change then there is specificity
What is effectiveness? What kind of designs are common in testing for this?
Does the treatment actually work in real-world clinical settings?
Studies in naturalistic settings
Quasi-experimental, uncontrolled designs are common
Can be conducted after efficacy research
What is treatment efficiency?
Cost effectiveness
Treamtents that cost the least are likely to be preferred if there is no great difference in outcome
Need to consider short-term and long-term gain (the more expensive therapy may be more effective in the long-term (cheaper in the long run))
What are some of the non-required but suggested factors concerning designation by the APA for ESTs?
Therapist training and monitoring
-need to know that the researcher was well trained and monitored
Consider investigator allegiance
-Any treatment tends to do better than other treatments when it’s conducted by people who are expert in its use
To deal with this, try and balance (e.g., if CBT is the focus of the study and IBT is the control used, have CBT experts conduct that therapy and IBT experts conduct that therapy
Comprehensive outcome
-Looking at general measures (quality of life, daily functioning) or, at least, looking at several measures (not just one specific diagnostic measure)
Follow-up studies:
-To see if gains are maintained
Clinical significance:
-Rather than only statistical significance (real-world change can be very different from statistical significance)
What are some problems/criticisms of ESTs?
RCTs are too rigid - inclusion/exclusion criteria are too strict
RCT designs place undue weight on efficacy rather than effectiveness
Selection of uncomplicated diagnostic groups fails to represent needs and responses of usual, complex, comorbid patients seen in practice
Not applicable to a diverse range of patients varying in personality, race, ethnicity, culture, etc.
Focus on brief, manuliazed treatment does not adequately or validly represent psychotherapy that is conventionally practiced
What is evidence-based practice? How is it different from ESTs? What are the benefits?
“…the integration of the best available research with clinical expertise in the context of patient characteristics, culture, and preferences.” (APA, 2006)
Going beyond the ESTs and looking at other types of research (including case studies, etc.), integrating with expertise and patient characteristics, etc.
More flexible and realistic approach than simple focus on ESTs
What is the three-legged stool? Who developed this idea? What does it relate to?
Best available research
Patient characteristics, culture, and preferences
Clinical expertise
Bauer, 2007
evidence-based practice
What are clinical practice guidelines?
a summary of scientific research (dealing with the diagnosis, assessment, and/or treatment of a disorder) designed to provide guidance to clinicians providing services to patients with the disorder.
The American Psychiatric Association, for example, has practice guidelines listed on its website that address the treatment of dementias, mood disorders, panic disorder, OCD, PTSD, borderline personality disorder, eating disorders, substance use disorders, and schizophrenia
What is the benchmarking strategy?
involves using the results of efficacy trials to form a standard (or a benchmark) against which the services provided to regular patients by regular clinicians can be compared
based on the benchmarking strategy, there is evidence from various countries that evidence-based treatments can be very effective when used in routine practice settings
11 questions to ask when critiquing the science of particular treatments
Was there a control group?
Was it an appropriate control group?
Was there random assignment to groups?
Would we be able to determine efficacy?
Is the treatment possibly efficacious or efficacious?
What would need to occur for this therapy to meet criteria for specificity?
Were the therapists trained adequately?
Who were the participants?
Were there enough study participants for the results to be meaningful?
Did the study evaluate adherence to treatment?
Did the investigators evaluate appropriate outcomes?
What is the difference between evidence-based and eminence-based practice?
Evidence-based: requires psychologists to be not only sensitive and empathic but also well informed about current research relevant to the services they provide. The effective scientist-practitioner thinks in a scientific manner and applies knowledge derived from research with care and compassion.
Eminence-based: practice based on tradition and authority; recommendations are accepted because the person delivering them is seen as an expert
What is the problem with eminence-based practice?
opinions of even recognized experts are just that—opinions—unless their views are supported by the best available empirical evidence
What are the 6 common errors in thinking in eminence-based practice?
Faulty Reasoning
False Dilemma
Golden Mean Fallacy
The Straw Person Argument
Affirming the Consequent
Appeal to Ignorance
What are:
-faulty reasoning
-false dilemma
-golden mean fallacy
-A form of argument that is inaccurate or misleading in some way. “Psychologists have provided effective services for decades without having research available on what makes treatment effective.”
-This fallacy takes the form of reducing the range of options available to just two (usually extreme) options. “Either I accept the treatment that the psychologist is suggesting, or I just give up trying to change.”
-This logical error involves assuming that the most valid conclusion to reach is a compromise of two competing positions. “I have heard that both cognitive and psychodynamic treatments can be helpful for the type of problems I have, so I really should look for a treatment that combines both cognitive and psychodynamic elements.”
What are:
-the straw person argument
-affirming the consequent
-appeal to ignorance
-This involves mischaracterizing a position in order to make it look absurd or unpalatable.
“Anyone who would prescribe a drug to treat my son’s symptoms just wants to turn active kids into zombies.”
-This logical error takes the following form: first, assume that x is a cause of y. Then, when y is observed, conclude that x must have caused it. “People who have schizophrenia always act in a bizarre manner. This person is acting bizarrely. So, obviously, this person has schizophrenia.”
-This mistake takes the form of arguing that, because there is no evidence to prove a position is wrong, the position must be correct. “There is no scientific evidence that having my patients sing and dance while they remember the trauma that they experienced harms them or is ineffective. So, of course, this new form of therapy has to be helpful.”
What is the difference between the deductive process and the inductive process in generating research ideas?
Deductive: the researcher uses a formal theory to generate a research idea
Inductive: for example, deriving an idea from repeated observations of everyday events. Even though the inductive process is not explicitly guided by theory, it is influenced by the researcher’s informal theories, including his or her theoretical orientation and general world view
What are the 5 steps to take in research after generating an idea?
First, the researcher conducts a systematic search of the published research on the phenomenon of interest.
Second, assuming that there is no research that has directly tested the idea, the researcher begins to formalize ideas so that they can be tested in a scientific manner - requires translating abstract ideas into something that can be measured (operationalizing an abstract concept)
Third, the researcher must carefully consider the extent to which the research idea may be based on cultural assumptions that may limit the applicability or relevance of the planned research
Fourth, the researcher must consider ethical issues in testing of the idea. For example, ethical considerations might make some research designs unsuitable, such as using random assignment in an experiment to determine the effects of violence
Finally, the researcher must draw together all the results of the previous steps to sketch out the study procedures. Along the way, some aspects of the planned study may need to be dropped or modified due to practical constraints (e.g., insufficient funds or a lack of appropriate measures).
What are some of the ethical considerations that must be taken in research?
participants may be vulnerable due to their psychological distress and/or may be receiving psychological services as part of the research
Psychologists are expected to apply the general principles (respect for the dignity of persons, responsible caring, integrity in relationships, and responsibility to society)
Prior to data collection, the researcher must obtain approval for conducting the research from the institution in which he or she works. In Canadian institutions, research ethics boards (REBs) ensure that the proposed research conforms to the Tri-Council Policy statement Ethical Conduct for Research Involving Humans
attention to the welfare of research participants (and animal subjects) and to honesty in the presentation of research findings
an ethical obligation to share their data with other researchers (once published)
What is the difference between internal validity and external validity?
internal: the extent to which the interpretations drawn from the results of a study can be justified and alternative interpretations can be reasonably ruled out
external: the extent to which the interpretations drawn from the results of a study can be generalized beyond the narrow boundaries of the specific study
What are 7 threats to internal validity?
History: influence of events that occur outside the context of the study
Maturation: Changes in the participants due to their psychological or physical development
Testing: Repeated testing may influence the results
Instrumentation: In longitudinal studies, changes in the definition of constructs and in their measurement
Statistical regression: Extreme scores on measures, both high and low, tend to be less extreme upon retesting
Selection biases: systematic differences in recruiting participants or assigning participants to experimental conditions
Attrition: loss of participants in a study over time
What are 5 threats to external validity?
Sample characteristics: the degree to which the characteristics of the research participants, such as their sociodemographic and psychological characteristics, map onto other samples and populations
Stimulus Characteristics and Settings: features of the study such as the institutional setting and the characteristics of those involved in the conduct of the study (e.g., therapists in a treatment study) may constrain the generalizability
Reactivity of Research Arrangements: By virtue of being in a study, participants may respond differently
Reactivity of Assessment: Participants’ awareness that their behaviours, moods, attitudes, etc. are being monitored
Timing of Measurement: The decision about when to measure variables may result in conclusions that are not true for all time points - may not be stable between measurement periods
What is useful about correlational designs and case studies?
once a research area is well developed, correlational designs are unlikely to add anything new to the scientific literature. On the other hand, in a relatively new research area, even a relatively simple case study may make a meaningful contribution to the literature
Designs such as these are often used in early research on a particular topic. As the knowledge on a topic evolves, there is a tendency for the research designs to become progressively more rigorous.
Typical case studies involve…
a detailed presentation of an individual patient, couple, or family illustrating some new or rare observation or treatment innovation
4 advantages of case studies
a valuable format for making preliminary connections between events, behaviours, and symptoms that have not been addressed in extant research
include a lot of detail (e.g., about treatment methods) making it easy to follow and replicate
can be a rich source of research hypotheses regarding the etiology or maintenance of disorders
can also be the initial testing ground for innovative assessment or intervention strategies
have heuristic value—that is, they draw the attention of other professionals to a phenomenon
3 disadvantages of case studies
do not allow for the rigorous testing of hypotheses
major weakness is that most threats to internal validity cannot be adequately addressed
alternative explanations cannot be ruled out in this simple research design
What is an A-B single case design?
A = period representing the level of symptoms prior to the intervention (also known as the baseline)
B = period representing the level of symptoms following the intervention
intervention effects are typically determined by visual inspection of graphed data
a number of statistical tests can be used to determine if statistically significant changes occurred
What is an A-B-A single case design?
similar to the A-B design except that the treatment is withdrawn after a few weeks and data continue to be collected for a second A period (i.e., a period in which no treatment occurs)
If symptoms return to pre-treatment levels, then a strong case can be made for the effectiveness of the intervention
What are the disadvantages of an A-B-A design?
it may not be possible to have the person refrain from using the treatment strategies during the second A period, especially if the strategies have been effective in reducing symptom levels for a few weeks
ethical considerations may make clinicians unwilling to remove, however briefly, a treatment strategy that appears to be working well
What are the most commonly used research designs in clinical psych?
correlational
What are some advantages of correlational designs?
can be analyzed with all types of statistics, including correlations, partial correlations, multiple regression, t-tests, or analysis of variance (ANOVA)
Come in many forms
Can be used to examine the underlying structure of a measure or a set of measures. This is known as factor analysis
Most frequently, correlational designs are employed to examine the relations among discrete variables in an effort to develop or test a conceptual model (e.g., moderator and mediator variables)
What is factor analysis? What is it used for?
a statistical procedure used to determine the conceptual dimensions or factors that underlie a set of variables, test items, or tests
can reveal which items “work” and which don’t
What are the two basic forms of factor analysis? What are they used for?
Exploratory factor analysis:
is used when the researcher has no prior hypotheses about the structure of the data
Confirmatory factor analysis:
is used to test a specific hypothesis regarding the nature of the factor structure
What is a moderator variable?
a variable that influences the strength of the relation between a predictor variable and a criterion variable
For example, the relation between the experience of stressful life events and psychological distress may be moderated by the type of coping strategies used
What is a mediator variable?
a variable that explains the mechanism by which a predictor variable influences a criterion variable
For example, the relation between parental psychopathology and child adjustment may be due, partially or entirely, to the quality of the relationship between parent and child.
What is structural equation modeling (SEM)?
What are the steps?
a comprehensive statistical procedure that involves testing all components of a theoretical model which combines elements of confirmatory factor analysis and mediator analyses
First, the researcher lays out a structural model that shows how all the variables in the model are related to each other - akin to mediator analyses
In a second step, the researcher considers how best to measure each variable in the model and selects multiple measures for each variable - akin to confirmatory factor analysis
What are the disadvantages of correlational designs?
a causal connection cannot be established, as the apparent effect of the first variable on the second could be due to the influence of an unmeasured third variable
experimental manipulation and random assignment to conditions are absent
What are the disadvantages of structural equation modeling (SEM) correlational designs?
SEM can determine only the extent to which a hypothesized causal model fits the study’s data—it cannot unequivocally demonstrate that the hypothesized causal model is true
SEM requires a relatively large sample (i.e., more than 200 participants)
What are the advantages of quasi-experimental designs?
involve some form of manipulation by the researcher (although no random assignment)
in many situations, it is simply not ethical or feasible to randomly assign participants to conditions so this design provides a compromise
The most frequently used quasi-experimental designs involve the comparison of two previously established groups of participants.
In the simplest design, one group receives the intervention, the other doesn’t. Data are collected after the intervention and then analyzed. This design is cost effective and relatively straightforward because only one wave of data collection is required.
