1
Q
Quantitative Research
A
- Seeks explanation or causation
- Objective, precise and measurable outcomes
- Numerical data is collected
- Surveys and questionnaires
- Biometrics (height, weight, bp, pulse, lab results, blood sugar, etc.)
- Data is analyzed using statistics
- Goal is to understand a phenomenon, through:
- Prediction, generalizability (findings from research can be applied in other areas, to other populations), causality (cause and effect)
- Provides research plan (to solving problems, answer questions, test hypotheses)
- Plans control (biases)
2
Q
confounding variables
A
Factors/variables that can cause change in the dependent variable that are not a result of the independent variable
Outside factors
3
Q
conceptual definition of variable
A
- derived from the literature
- this is how the variable is defined
- Eg. stress = increased heart rate and alertness
4
Q
operational definition of variable
A
- translates the conceptual definition into behaviours or verbalizations that can be measured
- this is how the variable is measured
- Eg. measure heart rate
5
Q
Quantitative: Testability
A
Measurable
Relationship proposed between X and Y
Is X related to Y?
What is the effect of X on Y?
6
Q
Theoretical (or Conceptual) Frameworks in Quantitative Research
A
- Makes the research more rigorous/strengthen quality of study
- A way to understand/explain the research being studied
- Provide an explanation of the relationship(s) between concepts in the study
(Cause and effect, Relationship, Difference) - Guides research (developing research question and hypothesis)
- Operationalizing variables –> how will we measure the variables (instruments?) (we have to measure variables the same way or else the results could be measured incorrectly → bias)
- Understanding results
- Goes beyond a “common sense” explanation
7
Q
research question
A
- Research study
- The research question and hypothesis lead to development of a study
- Can only be answered by conducting research in a study
8
Q
clinical question
A
- Consumer of Research (EIDM/EBP)
- A clinical question leads to the studies
- Can only be answered by findings in existing research
- Annotated bibliography assignment
9
Q
Quantitative: Characteristics of a Hypothesis
A
- relationship between variables is stated
- testable (measurable)
- hypothesis is either supported or unsupported (not proven)
- should state variables, population, and predicted outcomes
- theory based (should flow from research question, literature review to theoretical framework)
- Educated guess
10
Q
types of hypothesis
A
research hypothesis (alternative hypothesis) - Statement about expected relationship between variables (causal, directional, non-directional, difference)
statistical (null) hypothesis
- Predicts no relationship; there is no difference
- Rejection of null hypothesis is acceptance of research hypothesis
11
Q
variable relationships
A
- associative (there is a relationship)
- causal (one causes the other to change)
- non directional (there is an association)
- directional (same as causal) (one causes the other to change)
- null (no relationship)
- research (same as causal)
12
Q
critiquing hypothesis
A
Evaluate the wording for:
- Clarity of statement
- Implications for testability
- congruence/fit with theory/literature
- Appropriateness for research design used
13
Q
Quantitative Design Purpose
A
- Objectivity (literature theory)
- Accuracy (study flows from question)
- Feasibility (capacity) (is it possible to do this study?) (time, money, participants, ethics)
- Control (measures used to ensure study conditions remain the same to avoid bias on dependent variable p. 195)
- Control (ruling out extraneous/confounding variables)
- Homogenous sample (the characteristics of participants are similar between groups) (similar extraneous variables)
- Consistent data collection procedures
- Manipulation of the independent variable
- Randomization (equal chance of control or experimental group)
14
Q
internal validity
A
- Degree to which the experimental treatment, not an uncontrolled condition, resulted in the observed effects
- Confidence that you have that any change in dependant variable/outcome can be traced back to the independent variable
- Internal validity asks if it’s the independent variable (or something else) that caused or resulted in the change in the dependent variable
15
Q
threats to internal validity
A
- history (another specific event affected the DV eg. TV ads)
- maturation (development)
- testing (taking the same test repeatedly could influence outcomes)
- instrumentation (change in observation technique, using instruments properly)
- mortality/attrition (people dropping out of the study)
- selection (how participants are selected)
16
Q
external validity
A
Questions the conditions under which the findings can be generalized to the population; deals with the ability to generalize the findings outside the study
17
Q
threats to external validity
A
- selection effects (sample size, selection bias)
- reactive effects (Hawthorne effect - Idea that people behave a certain way when they know they are being studied)
- measurement effects (pre/post tests, instrumentation)
18
Q
sampling
A
- Sampling is a process of selecting a portion or subset of the designated population to represent the entire population.
- Sample selection based on inclusion and exclusion criteria
- A representative sample is one whose key characteristics closely approximate those of the population.
- Sampling strategies:
Non Probability—non-random
Probability—randomization of sample, more likely to be representative of population review
19
Q
sampling strategies - non probability
A
- convenience sampling (people that are most accessible)
- quota sampling (chose based on certain qualities, achieve a sample that represents certain quotas of the population)
- purposive (the researcher selects participants who are considered typical of the population)
- matching (Used to construct an equivalent comparison sample group by filling it with participants who are similar to each participant in another sample group (age, gender, education level, etc)
- network/snowball (For hard to reach populations (eg. drug users)) (A participant in the research helps you find other people to participate in the research)
20
Q
sampling strategies - Probability
A
- simple random (Everybody in the population has an equal chance of being selected) (Draw numbers randomly)
- stratified random (Population gets put into homogenous groups (everyone’s the same) and then people are randomly selected from the groups)
- multistage/cluster (selecting clusters/groups that already naturally occur and meet criteria)
- systematic (Randomly choose the first person and then every _th person is selected)
21
Q
quantitative research designs
A
- experimental
- quasi experimental
- non experimental/observational
22
Q
experimental designs
A
Testing interventions
Cause and effect relationships
Evaluating outcomes (efficacy and cost effectiveness)
Provide level II evidence
23
Q
experimental study properties
A
- randomization
- control
- manipulation
24
Q
inferring causality
A
1) The causal variable and effect variable must be associated with each other (make sure variables are associated (based on literature) before doing the RCT)
2) The cause must precede the effect
3) The relationship must not be explained by another variable (can’t be an alternative explanation based on another variable)
25
Types extraneous variables/confounding variables
- antecedent (occurs before the study. eg. age, gender, health status)
- intervening (condition that occurs during the study, eg. change in health status, developmental factors)
26
Solomon 4 Group Design
- Extra experimental and control group with no pretest
- Pretest questions may cue people to become motivated to partake in behaviors
- Could influence the post test
- Eg. how much did you exercise last week → people may become motivated to exercise because the survey made them realize they don't exercise enough
- So eliminating the pretest group can be compared with the pretest group so you can see if the results are the same for both
- Need a bigger sample size
27
experimental design advantages
Most appropriate for testing cause-and-effect relationships
| Provides highest level of evidence for single studies
28
experimental design disadvantages
- Not all research questions are amenable to experimental manipulation or randomization (difficult logistics in field settings) (ethical?)
- Subject mortality - subjects respond in a certain way when they are being studied
29
benefits of RCT's
- Most rigorous and robust research method of determining whether a cause-effect relations exists between an intervention and an outcome
- Reduce bias
- Rigorous is important - you can have poorly designed or implemented RCT’s
- RCT’s provide data for systematic reviews and meta analysis providing a solid base for synthesizing evidence from studies
30
how does RCT reduce bias
- randomization
| - blinding
31
Pilot RCT
- Assess feasibility, compliance, acceptability
- Assess the integrity of the study protocol
- Test data collection forms or questionnaires
- Randomization procedures
- Recruitment and consent
- Acceptability of the intervention
- Estimate of the mean and variability of primary outcomes for sample size calculation for larger trial
32
potential sources of Bias
- Participants - social desirability
- Researcher subjectivity
- Sample imbalances (eg. selection bias)
- Faulty data collection methods (Eg. measurement bias)
- Inadequate study design
- Flawed implementation
33
quasi experimental designs
- Testing interventions
- Evaluating outcomes (efficacy and cost effectiveness)
- Testing cause and effect relationships
- Provides level 3 evidence
34
Quasi-Experimental Design Features
- no randomization
- may not be a control (depends on nature of the variables)
- manipulation of IV
35
Quasi-experimental Design Types
- non equivalent control group design
- after-only non equivalent control group design
- time series design
36
After-only nonequivalent control group design
Similar to after only experimental design but no randomization
Two groups are assumed to be equivalent and comparable before the introduction of the independent variable
Commonly used for research conducted in field settings
Threats to internal validity
Are the comparison and experimental groups similar?
37
time series design
Useful for determining trends
No control group (threat to internal validity)
Bias and maturation (development) effects since the study is conducted over a longer period of time
Takes more resources and more time
Is more accurate → good design (multiple points of data collection)
38
Quasi Experimental Design Advantages
Practical and more feasible, especially in clinical settings
| Some generalizability
39
Quasi Experimental Design Disadvantages
Difficult to make clear cause and effect statements
| May not be able to randomize
40
Evaluation Research
- Use of scientific research methods and procedures to evaluate a program, treatment, practice, or policy
- Not a different design, researchers may use experimental or quasi experimental (even non experimental) designs to evaluate a program
- Not to develop new knowledge
- Are objectives being fulfilled and how?
- Determine reasons for success and failures?
- Direct course of experiment with techniques for its effectiveness
- Base further research
- Redefine the means to be used for attaining objectives and redefine sub goals
41
Types of Evaluation Research
formative
- A program is being assessed as it is being implemented
- Evaluation is focused on the process of a program rather than the outcome
Summative
- Outcomes of a program are assessed after completion of the initial program
42
Why Non Experimental Designs?
- Non experimental designs are done when you don't know a lot about a topic
- start developing knowledge
- If the non experimental designs finds a relationship then you may move up to a higher level design and test an intervention
43
why do studies use non experimental designs
- Construct a picture of a phenomenon at one point or over a period of time (cross sectional, longitudinal)
- Explore people, places, events, or situations as they naturally occur (birth weights, you can't control that it is naturally occurring - you can follow it)
- Researcher does not manipulate IV they study things that naturally occur
- Test relationships/associations and differences among variables
- Does not determine cause and effect!!
44
Non-Experimental Research Designs
Survey studies
Relationship or difference studies
- Correlational Studies
- Cohort Studies
45
kinds of survey studies
- descriptive
- exploratory
- comparative
46
Non-Experimental Design Advantages
- Difficulty explaining cause-and-effect relationships
- Only start to develop evidence of relationships
- Important to develop knowledge base on phenomenon of interest
- Useful in forecasting or making predictions
- Important designs when randomization, control, and manipulation are not appropriate or possible
- Useful in testing theoretical models of how variables work together in a group in a particular situation
47
survey research
- Survey research involves acquiring information about one or more groups of people by asking them questions and tabulating their answers
- Ultimate goal is to learn about a large population by surveying a sample of that population
- A face-to-face interview
- A telephone interview
- A written questionnaire
- Online questionnaires
48
survey disadvantages
- Captures a moment in time
- Relies on “self-reported” data (bias) (say what they think the researchers want to hear) (only remember positive/negatives)
- People’s descriptions of their attitudes and opinions are often constructed on the spot
- Participants sometimes misrepresent the facts in order to give a favourable impression
49
structured interview
the researcher asks a standard set of questions and nothing more, follows a strict script
50
semi structured interview
the researcher may follow the standard questions with one or more individually tailored questions to get clarification or probe a person’s reasoning (more flexibility in what/how the researcher asks questions)
51
advantages of face to face interviews
- Enables the researcher to establish rapport with potential participants and therefore gain their cooperation
- Thus, yield the highest response rates – the percentages of people agreeing to participate – in survey research
- People will share more when they feel they can relate to you
52
disadvantages of face to face interviews
- Time consuming (establish rapport, collect data)
| - Expensive
53
phone interview advantages
Less time-consuming
Less expensive
Potential access to virtually anyone on the planet who has a landline or cell phone
54
phone interview disadvantages
Response rate lower
Unable to establish rapport
Sample biased to those people without phones
55
Questionnaire Advantages
- Paper and pencil/online questionnaires can be sent to a large number of people
- Save the researcher travel expenses and postage is typically cheaper than long-distance telephone calls
- Participants can respond to questions with assurance that their responses won’t come back to them (privacy, anonymity)
- Participants may be more truthful
56
Questionnaire Disadvantages
- Majority of people who receive questionnaires don’t return them = low response rate
- People who do return them are not necessarily representative of the originally selected sample (bias)
- Participants responses will reflect their reading and writing skills and perhaps their misinterpretation of one or more questions
57
correlational design
- A correlational study examines the extent to which differences in one characteristic or variable are related to difference in one or more other characteristics or variables
- A correlation exists if, when one variable increases, another variable either increases or decreases in a somewhat predictable fashion
- These data are numbers that reflect specific measurements of the characteristics in question
58
caution with interpreting correlational results
- When two variables are correlated, researchers sometimes conclude that one of the variables must in some way influence the other
- In some instances, such an influence may exist
- Ultimately, we can never infer a cause-and-effect relationship on the basis of correlation alone
- Simply put, correlation does not, in and of itself, indicate causation
59
cohort design
- provides level 4 evidence
- These are the best method for determining the incidence and natural history of a condition
- The studies may be prospective or retrospective and sometimes two cohorts are compared.
60
prospective cohort design
- A group of people is chosen who do not have the outcome of interest (for example, myocardial infarction).
- The investigator then measures a variety of variables that might be relevant to the development of the condition.
- Over a period of time the people in the sample are observed to see whether they develop the outcome of interest (that is, myocardial infarction).
- In single cohort studies those people who do not develop the outcome of interest are used as internal controls.
- Where two cohorts are used, one group has been exposed to or treated with the agent of interest and the other has not, thereby acting as an external control.
61
retrospective cohort design
- These use data already collected for other purposes.
- The methodology is the same but the study is performed posthoc.
- The cohort is “followed up” retrospectively.
- The study period may be many years but the time to complete the study is only as long as it takes to collate and analyze the data.
62
advantages and disadvantages of cohort design
- The use of cohorts is often mandatory as a randomised controlled trial may be unethical; for example, you cannot deliberately expose people to cigarette smoke or asbestos.
- Thus research on risk factors relies heavily on cohort studies.
- As cohort studies measure potential causes before the outcome has occurred the study can demonstrate that these “causes” preceded the outcome, thereby avoiding the debate as to which is cause and which is effect.
- A further advantage is that a single study can examine various outcome variables. For example, cohort studies of smokers can simultaneously look at deaths from lung, cardiovascular, and cerebrovascular disease. This contrasts with case-control studies as they assess only one outcome variable (that is, whatever outcome the cases have entered the study with).
- Loss to follow up
- Retrospective studies less expensive as the data has already been collected
- Retrospective studies subject to recall bias
- Difficult to control for all other factors (confounding variables) that might differ between 2 groups
Nurs 1351
flashcards
Decks in class (3)
# Cards
