Week 3 — Research Articles & Quantitative Research Flashcards
Main sections of a research article include:
- Abstract
- Introduction
- Methods
- Results
- Discussion
- References
Abstract section should contain:
- Overview and purpose of the study
- General description and methods
- Highlights of results
- Statement of significance of results
- General conclusions
Introduction section should contain:
- Statement of the Problem
- Clinical Relevance
- Review of current literature
- Rationale and theoretical framework
- Specific purpose and hypotheses (guiding research
question)
often concludes with a research question… Although not
always present, specific research questions are often
delineated at the conclusion of the introduction section
Introduction section
Methods section should contain:
- Study design
- Criteria for and methods of subject selection
- Description and number of subjects
- Measurement methods and data collection techniques
- Data analysis procedures
Results section should contain:
- Narrative description of statistical outcomes
- Tables and figures that summarize findings
- Statements of support of the hypotheses or rejection of
the hypotheses
Discussion section should contain:
- Importance of the results
- Limitations of the study
- Suggestions for future research
- Applicability of the results to clinical practice
brief statement of the purpose of the study and it’s
findings
Conclusion section
Styled format of references?
APA
Research that is based around focus of study, research question, hypothesis, collected data, analyzed data, and interpreted data
Quantitative research
Measurement of outcomes using numerical data and
statistical data
Concerned with relationship between variables
Variables are the building blocks of the research
question
Quantitative research
3 general categories of quantitative research:
Experimental, descriptive, and mixed
This type of design is considered to be the most
rigorous. Includes the randomized control trial (RCT) which is considered to be the “gold standard”.
Experimental
What characteristics must be present to be considered as an experimental design?
- One sample
- One independent variable (IV)
- Participants must be randomly assigned to groups
- IV must be administered as planned
- Confounding variables are minimized & variables are
controlled by the researcher - One dependent variable (DV) on which all subjects
are measured - The hypothesis is attempting to show a cause & effect
relationship between the IV and the DV
The study sample consists of the people who are the
subjects in the research study and should be a “subset” of the population you are studying.
Representative sample
The primary traits of the target/representative population
Age, gender, sex, geographical location
Factors that would preclude someone from being a subject in a representative study:
Anyone considered confounding to the results
Types of variables
- Independent variables (active variables)- IV
- Dependent variables- DV
- Categorial variables
- Extraneous variables
predicts that the IV will have an effect on the DV and should attempt to show a cause & effect
relationship
Hypothesis
A variable that will predict or cause an outcome
independent variable
This variable is the response or outcome related to the effect of the IV.
dependent variable
variable that places individuals into categories
Categorical variable
When extraneous factors are recognized and kept constant so as to minimize their effects on the outcome
Control variable
Factors that can alter the dependent variable making it difficult to understand the relationship between the independent and the dependent variable
Extraneous variables
When extraneous variables are not controlled, they exert an influence on the IV and contaminate it in a way that could affect the study
Confounding variable
Blinding the members of the research team can help
Person conducting the randomization can be blinded to
the groups.
The person measuring the dependent variable (the
outcome) can be blinded from knowing which subjects
received which intervention. They would assess the
subjects to see if there were any changes, but they
wouldn’t know what caused the changes, if any were
present.
These are all ways to….
Minimize confounding variables
techniques used to reduce experimental bias by keeping the
subjects and/or the investigators ignorant of group assignments and research hypothesis
Blinding
2 types of blinding
Double-blind study
Single-blind study
Methods to improve validity and reliability
Pick appropriate outcome measures
Randomized groups
Administer IV as planned
Avoid fidelity problems
A special RCT in which clinical sites are randomly assigned to an experimental condition.
Cluster RTC
Advantages of cluster RCT
Prevention of bleeding or contamination
Balanced skill levels and possible biases
Disadvantages of cluster RCT
Complexity w/ training
Issues w/ fidelity and ethics
Loss of power in statistical analysis
Begins as a basic experimental design. Each participant
is randomly assigned to a group. Each group then
experiences each condition of the IV. Observations of
all groups are made following each intervention
Crossover
Crossover advantages
Type one errors reduce type two errors
Increase of statistical power w/ more measurements taken
Disadvantages of crossover
Sequence of carryover effect may make it difficult to determine superiority of intervention
With this method the researcher can gain insight as to how
long it takes for the intervention to cause an effect
Repeated measures design
This design is used to determine if one intervention can have multiple effects.
Multiple DV and one IV
Describes populations and is designed to document factors that describe characteristics, behaviors, and conditions of individuals and groups
Descriptive research
Two purposes of descriptive research
Explore observation
Better understanding of problem/phenomena
Steps in descriptive research
- OBSERVATIONS
- PATTERN NOTED
- HYPOTHESIS
- TESTING OF HYPOTHESIS
- GENERAL CONCLUSIONS
Types of descriptive research
Comparative
Developmental
Correlation
Survey
Retrospective
Measures behavior of 2 or more subjects
Comparative
designed to document how certain groups change over time on specific variables
Developmental
explores the relationship among variables without active manipulation of variables by the researcher
Correlational
used to collect data on subject characteristics, as part of descriptive, exploratory, or experimental studies
Survey
analyzes observations that were collected in the past
Retrospective
Design with no control group that usually leads to type 1 errors
Pretest-post test
when the researcher reports a relationship between the intervention and the outcome, when no relationship really exists
Type 1 error
Subjects in one naturally occurring group receive
one intervention and the subjects in a different
naturally occurring group receive a different
intervention
Non-randomized control group
▪ Subjects are not assigned to groups.
▪ In this type, the IV is a variable consisting of a pre-
existing condition
▪ This type of design is useful in answering
questions related to special characteristics of
disability groups.
Cross-sectional design
Involves collecting data over time using repeated
measures and completing a longitudinal analysis of the
data
Longitudinal research
All longitudinal research designs are characterized by
- They must be measured across time
- Observations must be made at more than two
points to identify the impact of time on the
trajectories
Types of longitudinal research designs
- Simultaneous Cross-Sectional Studies
- Trend Studies
- Time Series & Intervention Studies
- Panel Studies
An epidemiologic study in which the researchers select
groups of people based on whether or not they have the
disorder being studied
Case-control study
Disadvantages regarding a retrospective study
Unreliable data
A subject with a hypothesized risk factor is matched
with a subject who does not have this risk factor.
Multigroup cohort design
structured method of analyzing available studies to answer a clinical question
Systematic review
statistical technique for quantitatively combining the results of the multiple studies used in SR to measure the outcomes
Meta-analysis
Data is collected from subjects through structured
questions via phone, mail, or internet is known as
Survey research design
Research that is based on the performance of one subject
or a small group of people being considered as one
subject.
Single-subject research
Notation system for single-subject research design
A = baseline
B = intervention
C and other letters = additional interventions