6 HP 350 Lecture Slides Feb 23

6 HP 350 Lecture Slides Feb 23 - Chapter 8 HP 350 Spring...

Info iconThis preview shows pages 1–9. Sign up to view the full content.

View Full Document Right Arrow Icon
1 Chapter 8 HP 350 Spring 2010
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
2 Nonrandomized Designs For practical or ethical reasons, randomized experiments are not always possible. Nonrandomized studies are often referred to as Quasi-Experimental Designs. Like randomized experimental designs, these designs have treatment conditions, outcome measures, and sampling units. However, quasi-experimental designs do not randomly allocate sampling units to treatment conditions.
Background image of page 2
3 Third Variable Problem In randomized experiments, the process of assigning individuals to treatment and control groups by a random process helps to bolster claim of causality when X and Y covary. However, in nonrandomized experiments, a stronger possibility exists that X and Y covary due to another variable, referred to as the “third variable problem .” This third variable problem is a threat to internal validity for nonrandomized experiments.
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
4 Types of Nonrandomized Studies Nonequivalent-Groups Design Time-Series Design Single-Case Experimental Design Cross-Lagged Panel Design Longitudinal Design Cross-Sectional Design
Background image of page 4
5 Nonequivalent-Groups Design Program Group NR O X O Control Group NR O O NR   = Not randomized O    = Observation (Pre or Post test) X    = Treatment E.G.  Weight loss study- Allocate overweight children to  nutrition curriculum vs. control based upon existing groups. A between-subjects design with allocation to groups without randomization , i.e., “nonequivalent.”
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
6 Time-Series Design Defining feature is the study of some phenomena across time. Data is collected at several time points. Interrupted time-series design suggests an intervention or a naturally occurring situation, i.e., interruption, which separates time into “before” and “after” periods. No explicit control vs. treatment groups , but before period could be viewed as control and after period as treatment . Referred to as “found experiments” since naturally occurring. Diagram for Interrupted Time-Series: O O O O X O O O where O = observation, X = interruption E.G. Study of the murder rate in New Orleans between 2002 and 2008. Hurricane Katrina is the interruption in 2005.
Background image of page 6
7 Single-Case Experimental Design Single-Case Experimental Design Also referred to as a Small-N Experimental Research and N-of-1 Experimental Research . Considered a sub-category of interrupted time- series design. Includes the following characteristics: 1. One sampling unit or a few unites are studied. 2. Repeated measures are taken of the unit (a within-subjects design ). 3. Random assignment is rarely used.
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Illustrates a study on the effects of teachers’ attention on study behavior of a single student or single case. Single-Case Experimental Design
Background image of page 8
Image of page 9
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 06/10/2010 for the course HP 350 taught by Professor Lankenau during the Spring '09 term at USC.

Page1 / 41

6 HP 350 Lecture Slides Feb 23 - Chapter 8 HP 350 Spring...

This preview shows document pages 1 - 9. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online