Research Final Flashcards
EBP
External Scientific Evidence
Clinical Expertise/ Expert opinion
Client/ Patient/ Caregiver Perspectives
4 steps to EBP
- Forming PICA Questions
- Finding the evidence
- Assessing the evidence
- Making the decision
Interval validity
Internal: answers research question and provides evidence by controlling variance enough to provide a clear picture of the relationship btwn IV and DV
2 threats to internal Validity
Researcher Bias: reflexive about own voice/ perspective
Researcher Reactivity: designed account for influence of researcher on participants behavior
Results Section
Presentation of findings, free of investigator influence, no interpretation or study discussion. address hypotheses.
Measures of central tendency
mean, median and mode
Skewness
the lack of symmetry of a distribution
positively skewed, negatively skewed and symmetrical
Kurtosis
the general form and/or concentration of scores
Mesokurtic: bell-shaped/normal
Leptokurtic: extra peaked dist
Platykurtic: extra flat dist
5 basic types of qualitative research
Narrative, case-study, phenomenological, grounded theory and ethnographic
Narrative Research
collecting stories about individuals.
Case Study Research
in-depth description and analysis of a single individual
multiple sources of info
Phenomenological
focuses on a group of participants who share a life experience
determine presence of concept/idea/phenomenon
Grounded Theory
The goal is to develop a unifying theory that can explain a process, behavior, or interaction that is shaped by the phenomenology of participants
Ethnographic
a qualitative approach that is used when a researcher wants to describe, explain, and otherwise understand the perspectives of a group of individuals
4 ways narrative research is analyzed
Thematically (what was told)
Structurally (how it was told)
Contextually (where/when original action/storytelling took place)
Ideologically (according to specific worldview,mindset or cause)
Data Coding
Open - broad categories (domains), determine themes
Axial - several categories merged into 1 main theme.
Selective - fitting main themes into workable model to develop hypothesis
Parametric Statistics assumptions
1) The population parameter should be normally distributed
2) The level of measurement of the parameter in question should be interval or ratio
3) When there are two or more distributions of data to be analyzed, the variances of the data in the two different distributions should be about the same
4) The sample size should be large (e.g. 30)
Nonparametric statistics used when
assumptions for parametric aren’t met
1) The population parameter is not normally distributed
2) Nominal or ordinal data
3) Unequal variances
4) Small sample sizes (i.e. <10 per group)
Bivariate Descriptive Stats
examine association between 2 variables
valus from -1 to +1
Null Hypothesis
Ho.
States that there is no difference between the groups or no relationship between the variables
Active Hypothesis
The investigator’s best prediction of the differences and relationships based on evidence and theory
statistical analysis can ________ a hypothesis
disprove
Type 1 Error
researcher rejects null (Ho) hypothesis when it is in fact true
Type 2 Error
Research does not reject null (Ho) hypothesis when it is in fact false
alpha
0.05
level of significance
Inferential Statistics
allows researcher to assign meaning to results
between or within subject effects
ANOVA
Analysis of variance
compare differences when there are more than two groups for comparison and more than two conditions under which each group is tested. F ratio.
df
number of values in the final statistic that are free to vary
N-1
1, 2, and 3 way ANOVA and repeated measures ANOVA
1: 1 IV
2: 2 IVs
3: 3 IVs
repeated measures: allows 1 to make between-group comparisons for each group over time
effect size
- scale free, standardized
- provide an indication of the extent to which the value of a DV is explained by the IV
- provide an index of the plausibility of the null hypothesis
A priori
Allows one to determine the sample size required to reach a given alpha level
post hoc
Allows one to evaluate research that has been completed to determine if a failure to reject the null hypothesis was related to an insufficiently large sample size
discussion section
May supply further justification of the problem, choice of design, methodology, and further persuade the reader that the findings are meaningful. moves beyond details of current study. relates conclusions directly to problem. limitations of study.
3 categories of clinical data
Treatment data: during intervention
Generalization probe data: outside intervention
Control Probe data: behaviors or skills not expected to change as a result to intervention
Barriers to EBP
lack of tx outcomes research
employing hierarchies of evidence
RCT considered gold standard
Population
any group of individuals in which the researcher is ultimately interestded
Sample
subset of the population of interest
Sample size
usually we only look at a sample and generalize to the whole group by inference
Factors of sample size
purpose of the study, previous research, concerns about generalization, variability of attributes being investigated and research design
Selection criteria
Need sufficient description to replicate it
Are criteria for group composition clearly defined and defensible
Overlap between groups on the variables that distinguishes the group?
3 basic “ethical” principles
- respect for persons
- Beneficence
- Justice
IRB
Must submit these
Must gain approval of IRB before conducting research
Informed consent includes (3)
- what the study involves (purpose)
- description of materials, procedures, expected duration
- the risks
Privacy
an individual’s ability to control when and under what conditions others will have access to personal information.
Confidentiality
ability of other people to tie specific information or data to a given individual.
Nominal
Mutually exclusive labels with no “ranking” - e.g., pass/fail, male/female, group membership
Ordinal:
Mutually exclusive and ranking - e.g., mild, moderate and severe.
Interval:
All of the above and equivalence of units - e.g., standard scores on tests.
Ratio
All of the above and equivalence of units and a true zero point
Reliability
repeatability. The degree to which we can depend on the measure.
Criterion validity
empirical examination of how well the measure correlates with some outside validating criterion
Construct validity
empirical and rational examination of the degree to which the measure reflects some theoretical construct or explanation of the behavior or characteristic being measured.
Concurrent validity
two measures measure the same content (and construct).
Predictive validity
measure is used to predict some future behavior i.e. screening tools, GRE as a predictor for success in grad school.
Observer Bias
has to do with variability among individuals/researchers acting as judges or raters.
Non-interactional:
occurs when the researcher’s behavior affects recording of subject responses. Think of testing, too.
Interactional
interactions with the subject affect changes the subjects behavior
Data Analysis:
How will the data be organized, summarized, and assessed
Therapeutic Effect
A beneficial change as a result of treatment
Clinical Significance
A substantial therapeutic effect
Treatment Effectiveness:
Positive results expected of a clinical procedure when applied in practice.
