Working Research Ideas into Study Designs - PowerPoint PPT Presentation

Loading...

PPT – Working Research Ideas into Study Designs PowerPoint presentation | free to download - id: 6c29cb-M2RlN



Loading


The Adobe Flash plugin is needed to view this content

Get the plugin now

View by Category
About This Presentation
Title:

Working Research Ideas into Study Designs

Description:

Working Research Ideas into Study Designs General Steps to Designing a Workable Study Find a Problem Theorize About What You Think Is Happening/Missing Isolate a Few ... – PowerPoint PPT presentation

Number of Views:6
Avg rating:3.0/5.0
Slides: 34
Provided by: ElS46
Learn more at: http://wesamuels.org
Category:

less

Write a Comment
User Comments (0)
Transcript and Presenter's Notes

Title: Working Research Ideas into Study Designs


1
Working Research Ideas into Study Designs
2
General Steps to Designing a Workable Study
  1. Find a Problem
  2. Theorize About What You Think Is
    Happening/Missing
  3. Isolate a Few, Important Concepts
  4. Consider Ways of Measuring Your Variables
  5. Consider Problems Collecting and Interpreting
    Data
  6. Based on Interest and Feasibility, Choose the
    Best Ways to Measure Your Variables
  7. Sketch the Expected Relationship Between the
    Variables
  8. Reconsider Potential Problems with Interpreting
    Data

3
Find a Problem
  • E.g.
  • Something Youre Interested in
  • If a Problem per se, What May Be Causing It
  • Increased Tardiness, Criminality, Lowered
    Sociability (with Peers, Adults, Siblings, etc.)
  • Or Ways of Fixing It
  • School-Based, Community-Based, or Direct
    Interventions Increasing Opportunities Changing
    Background Consensus Modeling by Teacher, Peers,
    Literary Characters
  • Of How It Might Be Affecting Something Else

4
Find a Problem (cont.)
  • A Gap Noted in the Discussion Section of a Recent
    Article (e.g., 2009 or 2010)
  • Or A Topic You Can Find Little About in the
    Literature

5
Theorize About What You Think Is Happening/Missing
  • I.e., What Are the Underlying Causes or
    Mechanisms
  • What Can Prevent It or Influence It?
  • Guided by Previous Research
  • As Well as Your Intuition
  • N.B., You May Well Be
    Proven Wrong, But Often
    Thats When Things Are
    Most Interesting

6
Isolate a Few, Important Concepts
  • Usually Theorizing About What Is Going on Makes
    This Easier
  • Included Among the Concepts Can Be the Outcomes,
    Personal Characteristics, Social Factors, Etc.
    That Influence What Youre Interested in
  • These Will Be
    Your Variables

7
Consider Ways of Measuring Your Variables
  • E.g.,
  • Direct Observation Is Often Best
  • Usually Easiest Way to Be Objective and Valid
  • Number of Actions in Time Bins, Number of Ss
    Doing Something, Time Initiation (Reaction Time),
    Time to Completion

8
Consider Ways of Measuring Your Variables (cont.)
  • Self-Report Can Be Good
  • Of Course, Must Consider Intentional or
    Unintentional Report Biases

9
Consider Ways of Measuring Your Variables (cont.)
  • Previously-Collected Data
  • Grades, Census Data

10
Consider Ways of Measuring Your Variables (cont.)
  • Description of Others Behaviors
  • Usually Only When Nothing Else Is Feasible

11
Next, Consider Problems Collecting and
Interpreting Data
  • E.g.,
  • Honestly Noting Personal Biases
  • Confirming Stereotypes Without Looking Deeper
  • Foibles in Human Behavior/Perception
  • First Thoughts Lead Subsequent, See the Rest of
    the World as Similar to Our Own
  • Practical Constraints
  • IRB Approval, Time Constraints, Logistics

12
Based on Interest and Feasibility, Choose the
Best Ways to Measure Your Variables
13
Types of Measurements
  • Nominal
  • Levels of Variable Can Only Be Given Different
    Names
  • No Level Is More (Larger, Longer, Greater) Than
    Another
  • E.g., Gender, Race, Nationality

14
Types of Measurements (cont.)
  • Ordinal
  • Levels Can Be Put into Order Along a Dimension
  • But Cannot Say How Far Apart Levels Are
  • E.g., HS Rank, Likert-Scaled Responses, Top Ten
    Desserts, Letter Grades?

15
Types of Measurements (cont.)
  • Interval
  • Distance Between Levels Can Be Measured, But
    There Is No Zero-Point
  • E.g., IQ, Scores on Most Psychological
    Instruments, Temperature

16
Types of Measurements (cont.)
  • Ratio
  • Levels Can Be Measured
  • And There Is an Absolute Zero
  • E.g., Height, Weight, Age, Income
  • Interval and Ratio Data Are Often Considered as
    Dimensions

17
Sketch the Expected Relationship
Between the Variables
  • Pie Charts
  • Good for Percent/Proportion of Whole
  • Not Good for Much Else
  • Bar Graphs
  • Good for Simple Comparisons Between Groups
  • E.g., Nominal by Ratio Comparisons
  • Gender by Income

18
Sketch the Expected Relationship
Between the Variables (cont.)
  • Line Graphs
  • Usually Good for Comparisons Along Two Dimensions
  • E.g., Masculinity by Income
  • Scatter Plot
  • Essentially a Line Graph with All Data Points
    Showing

19
Pie Charts
20
(No Transcript)
21
(No Transcript)
22
Bar Graphs
23
Past Month Binge Alcohol Use and Illicit Drug Use
among Full-Time Employed Women Aged 18 to 64, by
Race/Ethnicity 2004 to 2008
24
Line Graphs
25
(No Transcript)
26
Scatter Plots
  • Gini index measure income inequality 0 equal
    distribution 100 one person has everything

27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30
Reconsider Potential Problems with Interpreting
Data
  • Repeat Process to Fix What You Can
  • Keep in Mind What You Cannot Fix

31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
About PowerShow.com