Ch 3: Science and Evidence-Based Practice Flashcards
A method at attempting to arrive at objective truths
Science
A proposed explanation for a problem or a set of observations
Hypothesis
The process of formulating explanations about the natural world and testing these explanations with experiments and data
Scientific Method
- The use of evidence to construct testable explanation and prediction to natural phenomena as well as the knowledge generated through this process
- knowledge about or study of the natural world based on facts learned through experiments and observation
Scientific Method
Three-pronged approach to working with clients, which consists of what?
- Making decisions based on the weight of the scientific evidence
- Field Observations
- Individual client needs and preferences
Three-pronged approach to working with clients
Evidence-Based Practice
Limitations of Field Observations
Biases and Errors, Uncontrolled Environment (Setting)
Steps for the Scientific Method
- Identify a problem or set of observations
- Formulate a Hypothesis
- Design a study to test the Hypothesis
- Collect, Synthesize and Interpret Data
- Discard or change Hypothesis if data does not support it or Continue testing data that supports the Hypothesis
The purpose for performing a Scientific Study
Identify a problem or set of observations
Scientist formulate a proposed explanation for a problem or set of Observations
Hypothesis
Scientist select the materials and methods for a Study
Design a study to test the Hypothesis
Scientists find that data either supports the hypothesis to varying degrees or does not support it
Collect, Synthesize and Interpret Data
An expected outcome generated from a hypothesis
Prediction
Hypothesis Testing Steps
Idea -> Generate Hypothesis -> Design Experiment -> Acquire/Analyze Data -> Accept or Reject
A Hypothesis or Set of Hypotheses for which a large body of high-quality evidence has been accumulated
Theory
An account of a person’s experience or event
Anecdote
A Variable in an experiment that a scientist makes no effort to manipulate or account for
Uncontrolled Variable
Top of the Hierarchy of Evidence #1
Systematic Reviews and Meta Analysis of Randomized Controlled Trials (RCT)
2 of the Hierarchy of Evidence
Randomized Controlled Trials (RCT)
3 of the Hierarchy of Evidence
Observational Research
4 of the Hierarchy of Evidence
Peer-Reviewed Editorials and Expert Opinion
Bottom of the Hierarchy of Evidence #5
Non-Peer-Reviewed Media, Including Anecdote and Tradition
News reports, Articles, Television Shows, Documentaries and Magazines
Non-Peer-Reviewed Media, Including Anecdote and Tradition
An opinion piece in a Scientific Journal; stance of scientists who have extensively studied a topic. Experts can help summarize existing scientific knowledge and how it is applied
Peer-Reviewed Editorials and Expert Opinion
Original research where scientists perform experiments and collect data - in contrast to Secondary research where scientists analyze data that has already been collected or published elsewhere
Primary Research
Research in which the researcher observes ongoing behaviors to determine correlation
Observational Research (often rely on self-reporting)
A relationship between 2 Variables
Correlation
Cause and Effect; Other Factors
Cause (Independent Variable) and Effect (Dependent Variable); Other Factors (Confounding Variables)
A type of scientific study/trial where participants are randomly assigned into different groups – one or more will be the intervention to be tested and one will be the control group. Groups are randomized and a control is used in an attempt to reduce potential bias in the trial.
Randomized-Controlled Trial (RCT)
- Establish Cause and Effect; greater control over confounding Variables
The variable that scientists manipulate in an experiment
Independent Variable
The ability to generalize the results of a study
External Validity
A review in which scientists Systematically gather all research on a topic and evaluate it based on predefined criteria and rules
Systematic Review
A Statistical Analysis of a group of studies to assess the overall weight of the evidence
Meta-Analysis
A meta-analysis is a statistical analysis of RCTs. It is a study of studies. Scientists gather RCTs that fit predefined criteria. They run statistics on the group of studies to gain an idea of where the overall weight of evidence lies.
Limitations of Meta-analysis
Flaws in study selection or analysis can impact results. For example, if the analysis includes poorly designed studies, it may make the conclusions unreliable
Scientific approach to gathering knowledge
Empirical
Scientific Attitude
Critical and Questioning
Scientific Method of Observation
Controlled
Scientific Method of Reporting
Unbiased
Scientific Concepts
Clearly Defined
Scientific Instruments
Accurate and Precise
Scientific Measurements
Reliable and Repeatable
Scientific Testability
Falsifiable
Based on Observation or Experience
Empirical; Knowledge grows as scientists collect data and study it in a systematic and objective manner.
The Ability to get similar results when something is measured under the same conditions
Test-Retest Reliability
- Measurements should be close in value on repeated measurements in succession by the same person. There are objective criteria for how measurements are made.
Allows scientists to narrow down what is true, by proving what is not true.
Process of Falsification
- allows scientists to narrow down what is true, by proving what is not true
Consistency of a Measure
Reliability
The Assessment of whether a tool is measuring what it is supposed to measure
Validity
A Sample of people under study is not representative of the larger population that scientists are looking to make inferences about
Selection Bias
Describes a study that looks backward in time
Retrospective
The inability to accurately remember past behaviors
Recall Bias
Observational Research Main Types
Descriptive (Describes individuals or populations)
and
Exploratory or Epidemiological Research (Looks for Relationships between variables in large populations)
Types of Descriptive Research
Case Studies
Surveys
Historical Research
Types of Exploratory or Epidemiological Research
Cross-Sectional Studies
Case-Control Studies
Cohort Studies
- Tracks a single person or group of people over time
- Discuss characteristics of or events that occur to a single person or group
Case Studies (Descriptive)
Primary Limitation of Case Studies
May not be Generalizable to larger populations
Questionnaires or interviews administered to samples of people
Surveys (Descriptive)
Primary Limitations to Surveys
They collect self-reported data, and can be subject to bias * Potential issues with the reliability, validity, or how wording of questions can influence people’s answers
Looks at events that have happened in the past and uses it to make inferences about the present and future
Historical Research (Descriptive)
Limitations of Historical Research
Information from the past may be incomplete or inaccessible * May be subject to the biases of the
researchers involved
Variables in a group at a specific point in time; a cross section of people
Cross-Sectional Studies (Exploratory or Epidemiological Research)
Most Useful for Population-based Surveys; to assess the Prevalence of diseases in a population
Cross-Sectional Studies (Exploratory or Epidemiological Research)
Primary Limitation of Cross-Sectional Studies
Impossible to assess Causal Relationships because it is a one-time “snap shot” measurement
Determine if exposure to something is associated with an outcome
* Scientists identify people with the outcome and similar people who don’t have the outcome
* Uses data that has already been collected or occurred in the past
Case-Controlled Studies (Exploratory or Epidemiological Research)
Group followed over time to determine association between an exposure and an outcome or disease
Cohort Study (Exploratory or Epidemiological Research)
- Cohort free of outcome or disease
- Researchers identify exposure or event of interest
- Cohort followed over time until disease or outcome occurs 4. Association between exposure and risk of outcome assessed
Cohort Study (Exploratory or Epidemiological Research)
Primary Limitations of Cohort Studies
Susceptibility to selection bias
* Subjects are followed for extended periods of time may withdraw or be lost follow-up
* If retrospective: less control over variables, and that subjects may be susceptible to recall bias
Scientists directly intervene with subjects and look at the impacts of the intervention
Interventional Designs
Interventional Designs
- True-Experimental Design: Randomized Control Trials (RCT)
– Gold Standard of research due to degree of rigor and control - Quasi-experimental Designs (Similar to RCTs, but subjects are not randomly assigned to groups)
True-Experimental Designs: Randomized Control Trials (RCT)
Parallel Design
Factorial Design
Crossover Design
Quasi-experimental Designs
Non-randomized Control Group
Time-Series
Researchers track two or more independent groups in parallel with each other
Parallel Design (True-Experimental Design: Randomized Control Trials (RCT)
Researchers test the impact of two or more variables simultaneously, rather than isolating a single variable
Factorial Design (True-Experimental Design: Randomized Control Trials (RCT)
The same group of subjects are assigned to both the treatment and control groups. Each subject acts as his/her own control.
Crossover Design (True-Experimental Design: Randomized Control Trials (RCT)
Strength of Crossover Design
A strength of crossover designs is that they reduce the impact of the variability between subjects (i.e., the differences in how each person responds to a particular diet).
Subjects have not been randomly assigned to that group. This carries more potential for bias than a RCT, because the control group may not be equivalent to the treatment group.
Non-randomized Control Group (Quasi-experimental Designs)
Set of data points that have been collected in time order.
Time-Series (Quasi-experimental Designs)
Strengths of Observational Designs
- Observational studies allow for the examination of large populations over long periods of time
- Help examine potential associations that occur over long periods of time, such as the association between diet quality and cancer risk
- Allow for the generation of hypotheses that can be further studied in RCTs
Limitations of Observational Designs
Observational studies cannot establish cause and effect
Correlation does not equal causation. The lack of control of variables increases the chance of false associations
Strengths of Interventional Designs
- The primary strength of interventional designs is that they allow for demonstration of cause and effect. Scientists can control many variables that would be impossible to control in an observational study
- Scientists can also explore mechanistic explanations for observed effects.
Limitations of Interventional Designs
- The high degree of control, while a strength, is also a limitation. Notably, the study may not reflect real-world conditions. This limits external validity (the ability to apply the results in practice with the general population).
- These designs are only capable of examining a limited number of variables at a time.
- Interventional designs tend to have small subject numbers and are generally performed for short durations.
The critical evaluation of a study by other Scientists
Peer-reviewed
Publications that compile Research Papers, Editorials and other academic writings related to a topic
Scientific Journals
Types of Peer-Reviewed Journal Articles
Position Stands and Scientific Consensus Statements
Systematic Reviews and Meta-Analyses
Individual Studies
Narrative Reviews
Thorough and valuable summations of the state of the evidence on various topics. They are usually written by large teams of scientists.
Position Stands and Scientific Consensus Statements
Scientists review all of the literature on a topic in a systematic fashion to assess weight of the evidence. A meta-analysis is where a group of studies are statistically analyzed to determine the weight of the evidence. These types of papers provide valuable overviews of the state of the evidence where position stands fall short.
Systematic Reviews and Meta-Analyses
Research is gathered in a more informal fashion compared to a systematic review and the authors give their assessment and interpretation of the overall weight of the evidence.
Narrative Reviews
High-Quality, Non-Peer Reviewed Sources of Information
Recently published academic textbooks
University public health websites (e.g., Harvard School of Public Health)
Academy of Nutrition and Dietetics website (https://www.eatright.org/)
U.S. government health websites (e.g., Centers for Disease Control and Prevention)
International health websites (e.g., Eat Well Guide or World Health Organization)
Conference presentations by primary researchers without a brand or ideology to defend or product or supplement to promote
Podcast interviews with primary researchers
