Tracey & Glidden-Tracey (1999) Integration of theory, research design, measurement, and analysis, Toward a reasoned argument Flashcards Preview

ECPY 767 (Final) > Tracey & Glidden-Tracey (1999) Integration of theory, research design, measurement, and analysis, Toward a reasoned argument > Flashcards

Flashcards in Tracey & Glidden-Tracey (1999) Integration of theory, research design, measurement, and analysis, Toward a reasoned argument Deck (7)
Loading flashcards...
0

Major problems with current research

1. Wrong question
a. Shift from “What test or measure or design should I use?” to “What do I wish to be able to say?”
2. Poor assumptive specification
a. lack of complete explication of all assumptions
b. poor accuracy of assumptions
c. conclusions do not follow logically from assumptions
3. Lack of integration of the components of research
a. Not discrete units, but integrated – got it?

1

Iterative (repeating a process) model – four components of research

1. Substantive theory
2. Design
3. Measurement
4. Analysis

2

Substantive theory

Explicate, delineate, etc.

3

Design

a. experimental v ex post facto
b. rigorous v relevant
c. exploratory v confirmatory

4

Measurement (dimensions of a measure)


a. concreteness – construct easily grasped by all?
b. type – discrete v continuous
c. perspective – self, parent, therapist
d. realism – naturalness, setting of measurement
e. level – local v global
f. biases – social desirability?
g. sensitivity – over the full variable range, or restricted?
h. ideographic v nomothetic – individuals v groups
i. dimensionality – bipolar, unidimensional or otherwise

5

Analysis

over reliance on null hypothesis testing; advocate use of factor analysis, cluster analysis, multidimensional scaling, and basic exploratory data

6

Main point of article?

Research process viewed as iterative, not linear; choices and assumptions in one component are interrelated with other components in the construction of a single coherent argument, each component needs to be revisited until the discrepancy among the assumptions in each of the components is minimal.