Week 1 & 2
Define pragmatism
Pragmatism is an approach that tells us that different methods can provide us with different important aspects of answering a research question (logical positivism and constructivism). Getting more than 1 perspective on a research question gives us a better understanding of the question as a whole
This is a bonus question from week 1 because I wanted another!
What is knowledge translation and what is what is one solution that has been proposed
Knowledge translation is a two way street between researchers and clinicians. Taking what researchers are finding in the lab and translation it into what clinicians are actually doing + Translation from experience using EBP from clinicians back to language researchers can understand
One proposed solution is knowledge to action. Clinicians and researchers working together. Researchers ask relevant questions, use clinically relevant protocols, and use action focused dissemination techniques to get the information out into the world in a way that is easily actionable
What is a quasi-experimental design and what are the pros and cons?
Bonus: what are the types?
Quasi experimental designs do not have randomization. Pros are that it mimics a clinical environment because you can't always control factors and it can address ethical issues because we can learn about inherent differences in our society and this can tell us how to plan the most beneficial interventions. Cons are that it may be hard to replicate the same results, can't make causal claims without randomization, confounding variables and bias.
Types: one group pre-test post-test, one group pre-test post-test repeated measure design, interrupted time series design (multiple pre and post for stabilization)
What are the three key features of analysis in single subject designs. define them. hint: what do we look at in the graphs
WHAT DID MADDIE THINK SHE WANTED TO DO BEFORE OT?
Level (change or no change)
Trend (accelerating vs. decelerating), (stable vs. variable), (linear vs. curvalinear)
Slope (shallow, medium, steep)
PHYSICIANS ASSISTANT
define and explain the difference between a prospective and a retrospective study
prospective: collect data over time after recruitment retrospective: collect data about time prior to recruitment
What is Evidence Based Practice and why is it important?
Combining the best available evidence, client perspectives, and your clinical reasoning.
It is important because you need to follow what the evidence says so we are actually improving outcomes for our clients
T/F: You would want to use quotation marks if you are searching for an exact phrase
WHAT IS MADDIE'S DUNCAN ORDER
True
ICED LATTE WITH ALMOND MILK: 1 PUMP CARAMEL, 1 PUMP BUTTER PECAN, 1 SPLENDA
What is an RCT experimental design. What are the interventions received (2 designs) and what are the 2 times of testing?
randomly assigned into intervention or control, able to make cause and effect claims on the outcome you saw.
crossover design(2 groups; pretest 1 and post test 1 then washout period to switch interventions pretest 2 and posttest 2): participants receive multiple interventions in a sequential manner. Each participant serves as their own control because they get the control and intervention. A washout period is included in between treatments to reduce carryover effects
parallel design(pre test post test): participants randomly assigned to control or intervention. Each group receives only one intervention (at the same time) and outcomes are compared between groups at the end of the study
types of tests: post-test only (only get data after intervention) pre test-post test (measure before intervention and after)
What are the three types of replication and explain them
direct: do the exact same thing
systematic: look for the exception to the rule, change 1 or 2 things
clinical: put it out into the world, translational version of replication, ex. see how teachers use the intervention in a real world classroom setting, still structure and measurements but with real world application
What is subject effect and experimenter effect?
subject effect is when the subject know they are in a study or being watched/recorded so they act differently
experimenter effect is when the researchers has unconscious actions that may influence the results of the participants (ex. may interact with groups differently)
What is a key difference in criteria between a systematic review and a meta-analysis?
WHAT IS MADDIE'S FAVORITE CUISINE?
Systematic reviews use quantitative synthesis (ex. plotting all effect sizes on the same graph) and meta-analysis use quantitative statistical synthesis/analysis (running a statistical analysis on the effect sizes)
MEXICAN FOOOOD
What is the difference between a research summary and a research synthesis.
Summary: Recap of the main points from a source, brief statement from a source, explains/describes the main idea, summary is a faithful representation of the author’s ideas
Synthesis: Themes identified and put together from multiple sources, a unified statement is created by combining separate materials, conveys your thoughts, takeaways (related to the topic at hand)
What are the three types of statistical analysis and explain them
per protocol: people that dropped out of the study are not included in the analysis.; treatment effect bias - can be a problem because it can show that the intervention is more effective than it is especially if people dropped out of the study because they felt like the intervention was not working for them (may not be an accurate reflection of what actually happened)
intention to treat: in intention to treat analysis, all participants are included in analysis from both the control and intervention group even if they did not do the whole intervention. Even if someone dropped out you just use all the data you have for them. Mimics the clinical setting because can't always complete the whole treatment. As risk for diluting the effects of the treatment because not everyone being analyzed as randomized into the intervention received the whole intervention
as treated: as treated includes the data from individuals who did not adhere to treatments (or dropouts) and analyzes their data based on the treatment they actually received. If someone was randomized into the intervention group but they drop out of the study, their data will be pushed and analyzed with the control group. This helps us understand a little more about how the group did as a whole. Potential for systematic bias if there is a common reason for dropout that is not being reflected in the data.
What is effect size in? What method is used to denote effect size in single subject designs? Explain the concept.
Effect size is the size of the difference or change. Not the same as statistical significance
"Non-overlap" is used to denote effect size in single subject designs. Non-overlap refers to the amount of data points that fall outside +/- 2 s.d. of the mean. Although it varies based off the type of research question you ask, generally non-overlap is good because it demonstrates that the intervention had a significant effect
What are the 5 things that you need to claim causality and explain them
strength of association: need statistics + statistical relationships
temporal sequence: cause can to come before effect
biological credibility: need to make sense biologically
dose response: more time sitting = more chance of ulcer, less time sitting = less chance of ulcer
consistency: needs to occur or happen in a lot of people (not just old, or men etc.)
What are the three different categories for classification of research and what is in each category
Methods: Quantitative, Qualitative, Mixed Methods
Design: Descriptive, Exploratory, Explanatory
Purpose: Basic, Applied, Translational
What is the parenthetical and narrative citation for this book in APA 7th edition:
Jackson, L. M. (2019). The psychology of prejudice: From attitudes to social action (2nd ed.). American Psychological Association. https://doi.org/10.1037/0000168-000
Parenthetical: (Jackson, 2019)
Narrative: Jackson (2019)
What are the 5 phases of clinical trials and what happens in each?
preclinical: mice/animal tissue, testing if the intervention works/helps/provides some sort of benefit
phase 1: focus is safety because first time being done in humans. secondary investigation of efficacy or mechanism. small group of healthy participants, not people from the protected class of research participants.
phase 2: Efficacy is more important factor, safety is still a big concern. Does this treatment actually slow the progression of MS. Start to use people with MS in the trails, safety is important because they may respond differently to treatment. Start to look at dosage at this stage.
Phase 3: generating enough data to move to regulatory approval of new therapies. Is this drug better than what we already have on the market (not is it better than nothing). Less side effects, works better etc.
Phase 4: better understand safety in a broader population or a specific subset of a population. Usually happens after the FDA approves something. Long term relevance and looking at the broad scope of people with MS (side effects 5 years down the road). Different age groups (ex. before looking at <50, now 50-75), before we didn't but now we are looking at people with comorbidities.
Explain the process of treatment for an N of 1 crossover design. How many participants are involved? What is the goal?
A single participant is involved
Pre test --> intervention A --> post test
WASH OUT
Pre test --> intervention B --> post test
goal: alternate between 2 potentially effective treatments repeatedly until 1 emerges as the best treatment for that individual participant
explain extraneous variable, confounding variable, and spurious relationship
give an example from lecture
WHAT IS MADDIE'S MIDDLE NAME
extraneous variable: may impact the dependent variable but not the independent variable
confounding variable: impacts the dependent and independent variable
spurious relationship: relationship that is a result of the confounding variable
Wriggly ex.
extraneous variable: wriggly feeling poorly
confounding variable: time of day
spurious relationship: Relationship of body temperature and wriggly cuddling is spurious and not causal ---> Time of day is influencing
LEE
List the three levels of theory in OT and explain each
Paradigm (OTPF + code of ethics): ethical code or domain of concern, outlines what to do but does not tell you "how" to do it
Models (MOHO, PEO): A simplified representation of structure and content that describes or explains relationships between concepts, allows for different lenses
Frames of Reference (Biomechanical, Rehabilitative): Most concrete level of theory, address specific areas of occupational disability and helps practitioners apply theory with individual clients in specific situations, tools and strategies about "how" to implement what you want to do
What are the 4 things that should be included in your search strategy?
how you searched (e.g. keywords and/or subjects)
search terms used (e.g. words and phrases)
search techniques used (e.g. nesting, truncation, etc.)
how you combined searches (e.g. AND / OR / NOT)
What are the types of randomization and explain each
WHAT IS MADDIE'S DOG'S NAME
simple randomization: truly random, computer program that puts half of participants in a standard treatment and half in an experimental group. Still a chance because it is so random that groups turn out not even in terms of demographics
block randomization: first group of people that sign up have a separate protocol, then next 20, then next 20. randomize into control and intervention from blocks instead of entire participant pool. Stops temporality effect from happening
stratified randomization: even more randomization. If you know a certain feature of your sample may play differently with the intervention. ex plasticity and age - put ages into strata (older and younger) and then randomize those groups individually to ensure you have people from both ages in both the intervention and control group
OAKLEY
What are the types of multiple baseline designs? List each and explain why it is an important method
across subjects: three different subjects do the same (baseline --> intervention). Important because it tells us more about responders vs. non-responders
across settings: same participants but do the intervention across multiple settings (ex. school and home). Different settings have different expectations so and this may lead to different behavior. Allows us to look at how settings impact performance.
across behaviors: look for different things when you do your intervention (ex. social interactions initiated vs. non-verbal prosocial behavior vs. appropriate ending of social interaction). This gives us a more multifaceted understanding of a phenomenon
list 3 kinds of bias and 3 kinda of non-probability sampling
Bias
measurement: leading questions, unaccessible questions (language), people know they are being measured and this may alter their responses
confirmation: researchers may structure the research in a way that will replicate what they already know about the world
recall: people don't have good recall/memory --> inaccurate data
Non-probability Sampling
Convenience sample: researcher uses the people that are most accessible to them
Snowball sample: researchers selects a few people and those people recruit even more people
Purposive sample: Participants are chosen because they possess particular traits, knowledge, or experiences relevant to the study