Research: Ch. 9

Time frames

when and how often with the data be collected.

T/F A true experiment requires that the researcher manipulate the independent variable by administering an experimental treatment (or intervention) to some subjects while withholding it from others.

A true experiment must have 3 things: random assignment, experiment/control group, & an intervention.

Controlling external factors

Achieving constancy of conditions Control over environment, setting, time Control over intervention via a formal protocol

T/F An experimental research design involves a nonrandomized controlled trial:

False!

Research Design and Time: Longitudinal Design

Data are collected two or more times over an extended period. Longitudinal designs are better at showing patterns of change and at clarifying whether a cause occurred before an effect (outcome).

Between-Subject Design

Different people are compared (e.g., men and women)

Non equivalent control group design

If pre-intervention data are gathered, then the comparability of the experimental and comparison groups at the start of the study can be examined. Nonequivalent control group before-after design Symbolic representation: O1 X O2 O1 O2 Without pre-intervention data, it is risky to assume the groups were similar at the outset. *Non equivalent control group after only design is much WEAKER. x 01 01 *involves the use of a comparison group that was not created through random assignment, and the collection of pretreatment data that permits an assessment of initial group equivalence.

Retrospective designs

In a retrospective design an outcome in the present is linked to a hypothesized cause occurring in the past . One example is a case-control design in which "cases" are compared to "controls" on prior potential causes .

Causality

Many quantitative research questions are about causes and effects. Research questions need to be addressed with appropriate designs.

Design is the best place to ________ _____- research, subjects, measurement, sampling

Minimize Bias

Experimental Condition

Must be designed with sufficient intensity and duration Attention must be paid to intervention fidelity (or treatment fidelity), that is, whether the treatment as planned was actually delivered and received.

Possible Control Group Conditions: Counterfactual

No intervention is used; control group gets no treatment at all. An alternative intervention is used (e.g., auditory vs. visual stimulation). A placebo or pseudo-intervention, presumed to have no therapeutic value, is used. Standard methods of care ("usual care") are used. A different dose or intensity of treatment or only portions of it are administered.

Descriptive Research

Not all research is cause probing. Some is descriptive e.g., ascertaining the prevalence of a health problem *Some is descriptive correlations

Biologic plausibility

The causal relationship should be consistent with evidence from basic physiologic studies. (additional criteria for causality)

Coherence

The evidence about the existence of a relationship should come from multiple sources. (additional criteria for causality)

Counterfactual

What would have happened to the same people exposed to a causal factor if they simultaneously were not exposed to the causal factor.

Relative timing

When will information on independent and dependent variables be collected—looking forward or backward in time? (forward is much stronger study)

Location

Where will study take place?

Characteristics of a True Experiment: Manipulation

The researcher does something to some subjects—introduces an intervention (or treatment). (true experiment)

Characteristics of a True Experiment: Control

The researcher introduces controls, including the use of a control group counterfactual. (true experiment)

Within-Subjects Quasi Experiments Time series design

gather preintervention and postintervention data over a longer period. Symbolic representation: O1 O2 O3 O4 X O5 O6 O7 O8 more data to compare.. see effects more over time. *No comparison group; information on the dependent variable is collected over a period of time before and after the treatment.

Summary:

-If the study does not say "experimental" it is most likely - quasi. They like to brag when they are a true experiment! -True experiment - those that provide us with the strongest evidence to use in our practice: *Intervention *Random assignment of subjects *Control/experimental group -Interpretations of causality only from experimental. For Quasi, they may infer, but should be made cautiously

Non experimental research

-If there is no intervention, the study is non-experimental (observational). -Not all independent variables ("causes") of interest to nurse researchers can be experimentally manipulated. *For example, gender cannot ever be manipulated. *Smoking cannot ethically be manipulated *Includes descriptive research studies that summarize the status of phenomena

Key Features of Quantitative Research Design -Is there an intervention, also known as a ______ ________

-Independent variable

Characteristics of a True Experiment

-Manipulation- the experimenter does something to some subjects- that is, there is some type of intervention -Control- the experimenter introduces controls into the study, including devising a good approximation of a counterfactual- usually a control group that does not receive the intervention -Randomization- the experimenter assigns subjects to a control or experimental condition on a random basis

Research questions that seek to illuminate ________ relationships need to be addressed with appropriate designs.

Causal

Many (if not most) quantitative research questions are about _______ & _______.

Cause & Effect

Correlational Designs

Cause-probing questions (prognosis or harm/etiology questions) for which manipulation is not possible are typically addressed with a correlational design -Define Correlation

Other Characteristics of a True Experiment: Randomization

Randomization (also called random assignment): The researcher assigns subjects to groups at random. -Typical assignment is to an experimental group or control group. -The purpose is to make the groups equal with regard to all other factors except receipt of the intervention.

The design is the basis for selecting the design- purpose and......?

Research question

Within-Subject Design

Same people compared at different times or under different conditions

Bias

The slanting of findings away from the truth Bias distorts the findings Research designs should be developed to reduce the likelihood of bias or to control for it

The Counterfactual Model of Causality: Def of Counterfactual

a counterfactual is what would have happened to the same people exposed to a "cause" if they simultaneously were NOT exposed to the cause.

Cohort

all of people who have experienced same event. Like everyone who has had an MI.

The Counterfactual Model of Causality: Def of Effect

an effect represents the difference between what actually did happen when exposed to the cause and what would happen with the counterfactual condition.

Possible Control Group Interventions: Attention Control

extra attention, but not the active ingredient of the intervention

Correlation

is an association (relationship) between variables and can be detected through statistical analysis. Correlation does not infer cause

Prospective designs

it's a correlational design, a potential cause in the present is linked to a hypothesized later outcome. This is called a cohort study by medical researchers. They are stronger than retrospective designs in supporting causal inferences—but neither is as strong as experimental designs.

Possible Control Group Interventions: Delayed Experiment

the intervention is given at a later date Symbolic representation: R O X O O R O O X O Where we think treatment will be most effective.

Descriptive Correlational

the purpose is to describe whether variables are related, without ascribing a cause-and-effect connection.

Control over confounds

through randomization

Within-Subjects Quasi Experiments One Group before-after designs

typically yield extremely weak evidence of causal relationships. Symbolic representation: O1 X O2

Panel study

we look at the same people, asking them the same questions over a period of time. ex. Ask about your eating habits one time, then 10 years later, reasking.

masking/blinding

withholding information to either one or all people involved.

Evaluation of Quasi-Experiments

-May be easier and more practical than true experiments, but: -They make it more difficult to infer causality -Usually there are several alternative rival hypotheses for results

Experimental Designs: Factorial Design

-More than one independent variables are manipulated simultaneously, allow researchers to test both main effects (effects from the experimentally manipulated variables) & interaction effects (effects resulting from combining the treatments) -Example: a 2 × 3 design = 2 levels of 1 independent variable, 3 levels of a second -Example: in a 2 × 2 × 2 design, 2 levels of 3 independent variables are manipulated

Experimental Designs: Pretest/post test (before-after) design

-Outcome data collected both at baseline and after the intervention; *Data collected both before introducing the intervention (at baseline) and after it. -Symbolic representation: R O X O R O O People may have learned from the pretest...

Experimental Designs: Post Test Design (or after only) design

-Outcome data collected only after the intervention; after random assignment & the intro of the treatment -Symbolic representation: R X O R O -R = Randomization; X = Receipt of intervention; O = Observation/measurement of dependent variable

Experimental Designs: Crossover Design

-Subjects are exposed to 2+ conditions in random order.*Are exposed to more the one experimental condition in random order and serve as their own controls; such as designs are within subjects designs. -Subjects serve as their own control. -Symbolic representation: R O XA O XB O R O XB O XA O Can get rid of other variables such as genetics, age etc. people aren't going to vary. A flaw would be the carry over effect. My thoughts and feelings from the first study may carry over into the other study

Additional Criteria for Causality were raised by Bradford-Hill in connection with the debate about evidence linking smoking to lung cancer. (2)

1. Biologic Plausibility 2. Coherence

Key Features of Quantitative Research Desing: Broad Design Options (3) Include:

1. Experimental (randomized control trial) 2. Quasi-experimental (controlled trial without randomization) 3. Non-experimental/Descriptive (observational study; no intervention) *Other options include 1. Within-subjects design 2. Between-subjects design

Key Design Features:

1. Intervention Key Question: Will there be an intervention? What specific design will be used? Design Options: Experimental RCT, quasi-experimental (controlled trial), non experimental (observational) design 2. What types of comparisons ill be made? Key Question: What type of comparisons will be made to illuminate key processes or relationships? Design Options: Within-subject design, between-subject design *Sometimes, the same people are compared under multiple conditions or at a different point in time (prep vs post op), this is called a within- subject design. Sometime different people are compared (e.g. those getting versus not getting an intervention); this is called a between-subjects design. 3. Control over confounding variables Key Question: How will confounding variables be controlled? Which confounding variables will be controlled? Design Options: Randomization, crossover, homogeneity matching statistical control

Criteria for Causality Three key criteria for making causal inferences:

1. The cause must precede the effect in time. 2. There must be a demonstrated empirical relationship between the cause and the effect. 3. The relationship between the presumed cause and effect cannot be explained by a third variable.

Key Design Features Cont'd

4. Masking Key Question: From whom will critical information be withheld to avert bias? Design Options: Open versus closed study; single-blind; double blind 5. Time Frames Key Question: How often will data be collected? When, relative to other events, will data be collected. Design Options: Cross-sectional (single point in time), longitudinal design (multiple points in time) 6. Relative Timing Key Question: When will information on independent and dependent variables be collected- looking backward or forward? Design Options: Retrospective (look back for potential causes or antecedents), prospective design (begin with a cause or antecedent and then see what outcomes ensue) 7. Location Key Question: Where will the study take place? Design Option: Single site versus multi-site; in the field versus controlled setting.

Which of the following would be a key criterion for causality: A. Cause occurring before the effect B. Third variable involved with cause & effect C. No empirical relationship b/w cause & effect D. Single-source evidence about the relationship

A. Cause occurring before the effect

Trend study

Ask same question, but not use the same people

Controlling intrinsic factors

Control over subject characteristics Methods: Randomization Subjects as own controls (crossover design) Homogeneity (restricting sample) Matching Statistical control (e.g., analysis of covariance)

Research Design and Time: Cross sectional Design

Data are collected at a single point in time.

Quasi Experiment

Involve an intervention but lack either randomization or control group Two main categories of quasi-experimental designs: 1. Nonequivalent control group designs Those getting the intervention are compared with a nonrandomized comparison group. 2. Within-subjects designs One group is studied before and after the intervention. May be easier and more practical than true experiments, but They make it more difficult to infer causality. *(controlled trials without randomization) involve manipulation but lack a comparison group or randomization. Strong quasi-experimental designs introduce controls to compensate for these missing components.

Designs (To Support Causal Inferences)

Support causal inferences. -Research designs vary in their ability to support the criteria for causal inference. -Experimental designs offer the strongest evidence for assuming cause and effect.

Design

The outline of the study or the overall apporach

Follow-up study

like if you had breast cancer, then they are asking many dif questions about what you are doing, may get blood drawn, do any family members have breast cancer, etc.