Basic Research Design

1
Basic Research Design

• Understanding and Using the Designing a Research
  Project Worksheet

2
Major Parts of a Research Project
1) Overall Research Question
2) Type of Design
3) Setting
4) Participants
5) Variables
6) Hypotheses and Research Questions
7) Statistical Testing
3
Article Used in This Lecture
Hungford, D. W. Williams, J. M., Furbee, P. M.,
Manley, W. G., Helmkamp, J. C., Horn, K., and
Pollock, D. A. (2003). Feasibility of screening
and intervention for alcohol problems among young
adults in the ED. American Journal of Emergency
Medicine, 21, 14-22.
4
Overall Research Question
What do you actually hope to learn during your
study?
5
Hungerford et al. study
Is it feasible to screen and perform a brief
alcohol intervention with young adults in the ED?
6
Type of Design

• Survey
• Epidemiological / Data Mining
• Randomized Study
• Nonrandom Study

7
Survey

• Person-2-Person
• Other Outlet
• Mail
• Web Based

8
Epidemiological / Data Mining

• Personally Collected Data
• Government Agency Data
• Hospital Records
• Other Source of Data

9
Ecological Study
Use data collected about populations and compare
them to other populations. No actual measure of
exposure is done.
10
Ecological Fallacy
Communities may differ in many factors, and one
or more of these may be the underlying reason for
difference in observed disease and death rates.
11
Randomized Study

• Non-Blinded Study
• Single Blinded Study
• Double Blinded Study

12
Nonrandomized Study

• Case-Control
• Cohort

13
Case Control Study
Start with people who have a specific outcome
(disease), and attempt to work backwards to find
out if there was exposure to a hazard of interest.
14
Cohort Study
Researcher observes health outcomes of a group
that has potential exposure to a hazard over a
period of time.
15
Hungerford et al. study

• Mixed Design
• Survey
• Non-randomized design
• -- Cohort Study

16
Types of Experimental Designs
17
Basics of Study Design

• O1 Observation 1
• O2 Observation 2
• X Study Intervention
• Group (Which research group does someone belong?)

18
Pre-experimental Designs
19
One-Shot Case Study Design
X ? O

• Expose someone to the intervention and then
  observe the outcome.
• You really do not know if your intervention
  causes the change.

20
One group pretest posttest design
O1 ? X ?O2

• Still a problem of unknown intervening variables
  (variables that cause the change that you are not
  measuring).

21
Static Group Design
Set Group 1 X ? O1 Set Group 2 ? O1

• Two set groups (EOH and OVMC) and all at EOH get
  the IV.
• You do not know if there are initial group
  differences.

22
True Experimental Designs
23
Pretest-Posttest Control Group Design
Group 1 O1 ? X ?O2 Group 2 O1 O2

• The outcome can be effected by exposure to the
  pretest when an IV is given.
• If you are given a test, then taught, and then
  given the same test Youll perform better
  because you knew what to look for during the
  lesson.

24
Posttest-Only Control Group Design
Group 1 O1 ? X ?O2 Group 2 O2

• It is more economical.
• Eliminates the possibility of pre- and posttest
  interactions.

25
Solomon Four-Group Design
Group 1 O1 ? X ?O2 Group 2 O1 O2 Group 3
X ?O2 Group 4 O2

• This design corrects for the flaws in the other
  designs.

26
Hungerford et al. study
This study was a One-shot Case Study. Everyone
who participated in the study received the
intervention.
27
Setting of Study
Where will the survey/study be conducted?
28
Hungerford et al. study

• Rural, University Based ED
• Study happened only between noon and 1 AM
  (for peak hours).
• School associated with the hospital has been
  ranked by The Princeton Review for being the 1
  party school.

29
Participants

• Will your participants have specific
  characteristics?
• How will you select/acquire your participants?
• Do you need a letter of consent?

30
Hungerford et al. study

• Will your participants have specific
  characteristics?

All participants had to be between the ages of 18
39. Only participants who had consumed alcohol
in the previous 12-months were included.
31
Hungerford et al. study
2) How are you going to select/acquire
participants?
Participants who did not speak English, were
combative, or were too drunk for participation
were excluded from the study. All participants
were recruited between the dates of August 1,
1998 to June 9, 1999.
32
Hungerford et al. study
3) Do you need a letter of consent for
participation?
YES!!! ?
33
Variables
Any entity that can take on different values.
34
Dependent Variables
The variable that is effected or not effected by
another variable in a research study.
35
Independent Variables
The variable that is being manipulated or
examined in a study to see if it effects the
dependent variable.
36
Example
In a study by Wrench and Booth-Butterfield
(2001), it was found that a physicians humor
orientation influenced her or his patients level
of satisfaction.
37
What was the IV and DV?
IV Physician Humor Orientation
DV Patient Satisfaction
38
Levels of Measurement
39
(No Transcript)
40
Hungerford et al. study

• Independent Variables
• -- Screen Positive or Screen Negative (aka were
  they heavy or non-heavy drinkers)
• -- Legal or illegal drinking age

41
Hungerford et al. study

• Dependent Variables
• Post intervention alcohol moderation
• Post intervention help-seeking behaviors
• ED staff perceptions of participant cooperation

42
Research Questions and Hypotheses
Research Questions (RQs) Questions that you
want specifically answered through your data.
Hypotheses (H) Predictions you make about data
associations in your study based on previous
research.
43
Activity
Based on the DVs and IVs from the Hungerfold et
al. study, can you identify any logical research
questions to ask?
44
Hungerford et al. study
RQ1 Will there be a difference between the
screen positive and negative participants and
their post intervention alcohol moderation
behavior?
45
RQ2 Will there be a difference between the legal
and illegal alcohol consumers and their post
intervention help-seeking behavior?
RQ3 Will the ED staff see the screen Positive
and Negative participants as being equally
cooperative?
RQ4 Will the ED staff see the legal and illegal
alcohol consumers as being equally cooperative?
46
H3 The ED staff will see the screen negative
participants as being more cooperative than
screen positive participants.
H4 The ED staff will see the legal alcohol
consumers as being more cooperative than the
illegal alcohol consumers.
47
Statistical Testing
Once you know what your research questions /
hypotheses and you know the level of measurement
of your IVs and DVs, the statistical
determination is very easy. See Picking a Test
Handout
48
Prepared by Jason S. Wrench, Ed. D. Medical
Educational Specialist West Virginia School of
Osteopathic Medicine