Research Design for Occupational Therapy – Flashcards

Unlock all answers in this set

Unlock answers
question
4 pillars of EBP
answer
1. Clients goals, values and circumstances 2. Research evidence 3. Information from the practice context 4. Clinical experience
question
5 steps in the process of EBP
answer
1. Convert info into a clinical question 2. Find evidence to answer the clinical question 3. Critically appraise the evidence 4. Integrate the evidence 5. evaluate the effectiveness of steps 1-4
question
3 main aspects of evidence
answer
1. Validity 2. Impact 3. Applicability
question
Pull information
answer
Answers the immediate clinical question with the best available e.g. clinical supervisor
question
Push information
answer
Develops professional knowledge for possible use in the future e.g. workshops
question
Types of questions
answer
1. Aetiology 2. Prognosis 3. Diagnosis 4. Intervention 5. Patients experiences and concerns
question
4 elements of PICO
answer
Population, Intervention, Comparison and Outcomes
question
Inductive research
answer
Investigating emerging and present phenomena to understand what is happening and why, from here we formulate the hypothesis
question
Deductive research
answer
Allows us to test the hypothesis to understand relationships and causality in our world
question
Randomised control trial
answer
Experimental study design where participants are randomly allocated to two or more groups that receive different interventions.
question
RCT advantages
answer
1. Suitable design to use if you want to trial all possible interventions 2. Controls against order effects on learning/satisfaction
question
RCT factors influencing findings
answer
1. Differences between groups at baseline 2. Changes in health during intervention 3. Measurement error 4. Research bias 5. Non-compliance with the intervention 6. Drop-outs
question
Cohort
answer
Observational longitudinal study, where one group of participants are followed over time
question
Case-Control
answer
Observational study involving 2 groups, cases have already developed the disease and are matches with other participants (controls) who are similar but have not developed the disease
question
Cross-sectional
answer
Sample a population at a particular point in time and measure them to see who has the outcome of interest
question
Quasi-experimental
answer
Experimental design that has no randomisation of participants to their interventions/control groups. Allocations is based on convenience groupings and can introduce sampling bias
question
Pre-post
answer
Experimental study design where measurements of the outcome of interest are taken at 2 points, before the intervention and after. Participants act as their own controls
question
Single subject design
answer
Experimental study design, where an individuals response to an intervention is measured over time. Measurements taken before, during, after and at follow-up
question
Correlation
answer
A descriptive study design where the researchers look for associations between 2 or more variables that are being measured
question
Case series
answer
A descriptive report on a series of clients (cases) who have an outcome of interest or have received the intervention being studied. There is no control group.
question
Level 1 level of evidence
answer
Systematic review of RCT
question
level 2 level of evidence
answer
Randomised controlled trials
question
Level 3 level of evidence
answer
Quasi-experimental, pre-post, cohort, case control
question
Level 4 level of evidence
answer
Correlation studies of multiple sites
question
Level 5 level of evidence
answer
Correlation, descriptive, qualitative studies and experts opinion
question
Nominal data
answer
Name the categories e.g. gender
question
Ordinal data
answer
Categories are ranked in order e.g. likert scale
question
Interval data
answer
Numeral data that can hold any value e.g. temperature
question
Ratio data
answer
Values can be compared as multiples of one another. There is a true 0 that has meaning e.g. I can go twice as fast as you
question
Parametric data
answer
- Normally distributed - t-score - Reported using a mean and standard deviation - Continuous data (interval and ratio)
question
Non-parametric data
answer
- Not normally distributes - z-score - Data is reported using median and interquartile range - Positively/negatively skewed - Categorical data (nominal and ordinal)
question
Intervention-to-treat (ITT)
answer
Conservative approach - Data are analysed to original random allocation to intervention or control (the way the participants were intended to be treated)
question
Hypothesis
answer
An educated guess that predicts outcomes of a study within a certain population
question
Inductive Hypothesis
answer
Based on trends or patterns seen in clinical settings
question
Deductive hypothesis
answer
Based on a theoretical premise and predicts what may occur
question
Null hypothesis
answer
No difference, we cannot accept a null hypothesis only reject or do not reject
question
Alternate hypothesis
answer
Predicts that any observed difference between the means of 2 groups is not due to chance
question
Testing a hypothesis
answer
Efficacy and effectiveness
question
Type 1 error
answer
- Saying there is a difference when there really isn't - We reject the null hypothesis when it is in fact true - We have said a real difference exists but the differences are due to chance - Denoted by the alpha.
question
Type 2 error
answer
- Saying there is no difference when there really is - We do not reject the null hypothesis when it is in fact false - We have said there is no real difference but the differences are real and are not due to chance - Denoted by the beta
question
Statistical power
answer
- When determining the chance of making a type 2 error - Denoted by 1-B
question
Statistical power is concerned with 4 factors
answer
1. Level of significance 2. Sample size 3. Sample variance 4. Effect size
question
Variance of the data
answer
- Power increases as variance reduces - When variability within a group is large, differences between groups will be less obvious - Variance can be reduced when sample size is increased
question
Effect size
answer
Measure of the degree to which the null hypothesis if false. It tells us how big the difference between 2 groups is. - 0 = little effect of intervention - 1 = large effect of intervention
question
Sample size
answer
The larger the sample the greater the statistical power
question
Factors influencing sample size
answer
- Finding and recruiting eligible participants can be challenging - Eligible participants might not consent - Clinicians may forget to enrol eligible participants - Dropouts - Allow for attrition
question
Independent (unpaired) t-test
answer
- Parametric test - Comparison between 2 independent data sets
question
Paired t-test
answer
- Parametric test - Used in repeated measures to improve control over extraneous variables - two related scores e.g. data from time 1 and time 2
question
Analysis of Variance (ANOVA)
answer
- Parametric test - When we have 3 or more samples to test the null hypothesis of no differences
question
One-way ANOVA
answer
- Parametric test - When there is only one independent variable
question
Repeated Measures of analysis of variance (RANOVA)
answer
- Parametric test - Comparing one group that is exposed to different interventions (3 or more related scores)
question
Mann-Whitney U Test
answer
- Non-parametric test - To determine differences in the mean scores between 2 sets of data that came from independent groups
question
Wilcoxin Signed-ranks test
answer
- Non parametric test - Comparing 2 related scores
question
Kruskal-Wallis analysis of variance by ranks
answer
- Non parametric test - Comparing 3 independent groups
question
Friedman two-way analysis of variance by ranks
answer
- Non-parametric test - Comparing 3 related scores
question
Chi-square statistic
answer
- Parametric test - Nominal data (whats your preference) - Compares expected frequencies and observed frequencies - E.g. cats vs dogs
question
Internal Validity
answer
Reflects the degree to which a conclusion about causal relationship between variables in a study can be properly demonstrated
question
Threats to internal validity
answer
- History - Confounding - Maturation - Natural progression - Attrition - Testing effects - Instrumentation effects - Regression to the mean - Selection bias - Experimenter bias
question
External validity
answer
The extent to which the results of a study can be generalised to other situations, to other people and to other time periods
question
Threats to external validity
answer
- Sample is small - Sample has characteristics that are not representative to the wider population - Excluding participants who are not compliant with the intervention - Study is conducted in a geographical or social environment that is specific - Whether results can be generalised to different periods of time in the past/future
question
Univariate
answer
data collected on a single variable and then characterised with descriptive studies
question
Types of descriptive research
answer
1. Cross-sectional 2. Case study 3. Cohort 4. Case-control (retrospective) 5. Correlation
question
Cohort advantages
answer
- Able to determine the incidence of the outcome of interest - Can establish a time sequence
question
Cohort disadvantages
answer
- Not useful for studying disorders that are rare - Prone to attrition (people drop out over time)
question
Case-control advantages
answer
- Don't have to wait years to find out what will happen to find a cause-effect relationship - Can look at exposure/non-exposure to treatment or risk factors
question
Case-control disadvantages
answer
- The 2 groups may not be well matched - Reliant on people's self-reporting of their history - Bias in the way that questions are asked - Reliant on archival medical records, which may be lost of incomplete
question
Correlation Coefficients
answer
Used to quantitatively describe the strength and direction of the relationship between 2 variables (r)
question
Spurious relationship
answer
2 or more events are independent of each other but they happen at the same time by coincidence or because of an unseen third variables
question
Pearson correlation
answer
- Tests associations between 2 variables - Used for parametric data - normally distributes - Continuous/ratio data - 0 = no relationship - 0.75+ = strong to excellent
question
3 aspects to consider when reporting associations between variables
answer
1. What is the direction of the relationship? 2. What is the strength of the relationship? 3. Is the relationship significant?
question
Spearmans Rho (Sr)
answer
- Non-parametric data - Examines the difference between the 2 sets of rankings for each participants - Scores are on an ordinal scale
question
Prevalence
answer
Is a measure of how much a health condition exists in a population at a point in time
question
Incidence
answer
Is a measure of the rate of new cases of the health condition in a specified time period
question
Probability
answer
The chance of an event happening
question
Absolute risk
answer
Refers to the incidence or natural frequency of the disease in the population
question
Absolute Risk Reduction (ARR)
answer
Indicates the actual difference in risk between those who have received an intervention and those who have not = Control event rate (CER) - Experimental event rate (EER)
question
Number needed to treat
answer
Identifies how many patients would need to be treated before a benefit can be seen. It describes the effectiveness of an intervention in terms of patient numbers
question
Relative Risk
answer
The likelihood that someone who has been exposed to a risk factor will develop the disease compared to someone who has not been exposed
question
Relative Risk Reduction (RRR)
answer
We consider a decrease in risk/increase in benefit that is associated with an intervention = (CER-EER) / (CER*100)
question
Cohen's D
answer
- Magnitude of difference between 2 compared groups - What was the effect of the intervention - Parametric test - Continuous/ratio data - 0 = hardly any difference - 1+ = very large difference between 2 data sets
question
Systematic Review
answer
A rigorous process of searching, appraising and summarising evidence. Addresses the effectiveness of an intervention and looks at the accuracy of diagnostic tools.
question
Characteristics of a Systematic Review
answer
- Clear aims and relevant criteria - Transparent and reproducible methods - Rigorous search designed to locate all eligible studies - An assessment of validity of the findings - Systematic presentation and synthesis of the included studies
question
Systematic review process
answer
- State the study object - Develop the protocol - Develop a search strategy - Conduct the search - Retrieve relevant papers - Screen and select papers that meet criteria - Evaluate methodological quality of studies - Critical appraisal - Analyse, synthesise and report findings
question
Inclusion/exclusion criteria based on
answer
1. Types of studies (level of evidence) 2. Partipants 3. interventions 4. Outcomes
question
Measurement error
answer
- Random error, systematic error - Calculated by: true score = observed test score - (systematic error + random error)
question
Random error
answer
- Environment (temperature/background noise) - Client (feeling sick) - Therapist (recorded wrong test score)
question
Systematic error
answer
- Outcome measure (unreliable or non-standardised) - Therapist (doesn't follow clinical guidelines)
question
Ways to test validity
answer
1. Ask the experts (face) 2. Cover all aspects (Content) 3. Compare to an external standard (criterion) 4. Compare to an idea (construct)
question
Face validity
answer
The extent to which a test appears to measure what it is supposed to measure
question
Content validity
answer
The relevance of a test to the construct that it says it measures
question
Criterion-related validity
answer
How its results compare with the results from another relevant measure (concurrent and predicted)
question
Construct validity
answer
Refers to how valid a test or assessment is in terms of the validity of the construct (the underlying theory or model) that the test says it is measuring
question
Convergent validity
answer
Get the same answer to what you expect
question
Discriminant validity
answer
Get a different answer to what you expect
question
Reliability
answer
A measure that gives us the same result over and over again in the absence of any real change (repeatability or consistency)
question
Inter-rater reliability
answer
Used to assess the degree to which different raters give consistent estimates of the same population - Nominal = Cohens Kappa
question
Test-retest reliability
answer
Used to assess the consistency of a measure from one time to another - Non-parametric ordinal = spearmans rank - parametric ratio = person correlation
question
Internal consistency reliability
answer
Used to assess the consistency of results across items within a test and the total score - Ordinal = Cronbach's alpha
question
Standard error of measurement (SEM)
answer
Gives an indication of the dispersion of the measurement errors when you are trying to estimate a clients true scores from their observed assessment scores
question
Linearity
answer
How easy it is to change the score in different parts of the scale
question
Clinical utility
answer
How easily can the outcome measure be used in clinical practice? (cost, training, time taken, equipment used)
question
Raw score
answer
The score obtained from a specific testing procedure
question
Standard score
answer
Developed so that the score can be used to understand performance of people from different groups or across sub-tests
question
Derived score
answer
A raw score that has been converted to a second scale (confidence intervals, percentile scores)
question
Percentiles
answer
Derived score that indicates how many people score above and below the percentile score
question
Confidence interval
answer
A way of calculating measurement error, tells you how confident you can be that the true score is within a given range
question
Qualitative research
answer
Allows the researcher to understand the perspectives of the people they engage with: this facilitates a client-centred approach to professional practice
question
Characteristics of qualitative research
answer
- Inductive - Occurs in naturalistic settings - Uses purposive and convenience sampling - Data consists of descriptions that are analysed using both qualitative and non-qualitative means - Sources of data include: texts, fields and action processes
question
Sampling for qualitative research
answer
- Include participants who have knowledge in the area being measured - Purposive and convenience sampling - No requirement for a particular size - Saturation point
question
Data gathering strategies for qualitative research
answer
1. Participation in the setting 2. Observation 3. In-depth interviews 4. Analysis of documents and objects
question
Phenomenology
answer
Aims to describe experiences from the perspective of people who have the experience (e.g. people who have cancer). Only those who have lived experience are fully knowledgable
question
Ethnography
answer
Aims to understand the underlying patterns of behaviour and meaning in culture
question
Steps of ethnography
answer
1. Gain access to culture of interest 2. Collect data (observation/interview) 3. Record data in log book, diary or audio 4. Analysis of data from description to explanation to reveal meanings and generate a theory 5. Impressions are verified with the informants and reported once findings are determined as accurate
question
Grounded theory
answer
Purpose is to construct a theory from qualitative data and the theory emerges from the experiences of the people interviewed. Sets of categories emerge that are used to make a generalisation
question
EBP is
answer
- A process of looking for understanding - Associated with research - A potential threat to the OT's practice
question
Participation Action Research (PAR)
answer
Generation and ownership of knowledge is associate with power and influence. People have capacity to work towards the solution of their own problems
question
PAR elements
answer
- Joint ownership of the research - Collaborative conduct of the research - Immersion of the subjects being studied in every stage of research
question
PAR processes
answer
- Collective investigation of problems and issues - Collective data gathering and analysis - Collective action and evaluation of outcomes
question
Concerns with PAR
answer
- Tokenism - Time frames - Cost - Ownership of knowledge
question
Q Methodology
answer
A research method used to understand peoples subjectivity. To explore and generate a better understanding of people's perspectives and beliefs
question
Steps in conducting Q Methodology
answer
1. Developing the Q-sort pack - develop statements 2. Administering the Q-sort - sort the statements 3. Factor analysing the data - find factors that summarise the pattern 4. Interpreting factors - interpret the main factors that have emerged
question
Q Methodology Advantages
answer
- Allows people to express their viewpoints - Rare missing data - Can use with different populations - Require small sample of participants
question
Q Methodology Disadvantages
answer
- Time consuming in administration - Comprehensive instructions - Not designed to show how many people have a particular view - May not always represent broader population
question
Steps in data management
answer
- Tape record the conducted interview - Transcribe the recorded interviews - Write up observations - Transcribe the spoken words - Write comments and explanations of what is going on
question
Analysing the data
answer
1. Free-form margin notes 2. Conceptual labels 3. Index codes with operational definitions 4. Audit trail
question
Credibility
answer
- Trustworthiness of data - Related to the true picture of the phenomenon - Triangulation, member checks and time sampling
question
Transferability
answer
- Trustworthiness of data - Cn the findings be transferred to other situations or can the participants be compared to the population of interest? - Purposive sampling
question
Dependability
answer
- Trustworthiness of data - Consistency between the data and findings - Audit trail, code-recode strategy, triangulation, peer examination
question
Confirmability
answer
- Trustworthiness of data - Have steps been taken to limit the bias in the research? - Triangulation, reflexivity (journalling), peer review
question
Triangulation
answer
The aim is to enhance validity by using more than one method, researcher, data set and theories
question
Journalling
answer
To document assumptions, beliefs, personal feelings related to the study, the participants or the data. This reflexivity separates the researchers subjective opinion from the data analysis
question
Member checks
answer
Data is gathered, recorded, transcribes and analysed and transcripts are given to participants to ensure accuracy
question
Coding strategies
answer
Essentially looking for certain words, phrases, patterns of behaviour or ways of thinking
question
Types of naturalistic enquiry
answer
- Phenomenology - Ethnography - Grounded theory - Action research (PAR)
question
Human research
answer
- Taking part in surveys, interviews and focus groups - Undergoing psychological, physiological or medical testing - Being observed by researchers - Access to their information
question
Beneficence
answer
Minimise risk to the greatest extent possible and maximise the potential for benefits to be gained from the research outcomes
question
Respect for human beings
answer
- Justice - Access to treatment - Storage, access and disposal of personal information
question
Informed consent must state
answer
1. Purpose of the study 2. What participation involves 3. The benefits and risks of the study 4. Voluntary participation 5. Withdrawal 6. confidentiality of information
question
Vulnerable participants
answer
- Pregnant women - Children, young people - People in dependent/unequal relationships - People unable to give consent (Seeking medical care) - People with cognitive/intellectual disability or mental illness - Aboriginal or Torres Strait Islanders
question
Ethics in publishing research
answer
- Informed consent - Approval from an independent ethics committee - Disclose conflict of interest - The role of editorial boards for journals
question
Evidence Based Practice
answer
1. Having a clinical question to answer 2. Seeking sources of credible information 3. Accessing the information 4. Critically reading the information 5. Evaluating the usefulness of the info 6. Trialling it in your practice 7. Evaluating the effect of the info
question
Barriers to implementing evidence
answer
1. Attitudinal barriers (people who are resistant to change) 2. Knowledge and skills/confidence barriers 3. Social influences (expectations of clients)
question
Strategies to overcome barriers of evidence
answer
- Educational materials - Educational meetings - Educational outreach visits - Reminders - Audit (summary of clinical performance over time) and feedback
question
Sensitivity
answer
is a technique used to determine how different values of an independent variable impact a particular dependent variable under a given set of assumptions.
question
Objectivity
answer
Assumes that a truth or independent reality exist outside of any investigation or observation.
Get an explanation on any task
Get unstuck with the help of our AI assistant in seconds
New