lis 570 how can i measure that? selecting variables. developing indicators for concepts
TRANSCRIPT
LIS 570
How can I measure that? Selecting Variables. Developing indicators for
concepts
Summary
Define concepts Identify dimensions Identify variables Identify indicators Evaluate the indicators
Moving from the abstract and theoretical to the empirical and measurable
Descending the Ladder of Abstraction
Problem statement Statement of Concepts
AbstractionsNot empirical
Process of making empirical Operationalization (Bouma) Clarifying concepts (De Vaus)
Descending the Ladder of Abstraction
Process Develop a conceptual definition for the concept(s)
locate a range of definitions select one for the study
Find variables for the concept
Assess validity of variables
Develop a Conceptual Definition
Conceptual definition “Dictionary type definition” Consists of more concepts
Leisure Activities of Interest“activities-apart from obligation to work, family & society to which a person turns at will” “interests = those activities which people actively seek out”
Locate and select a definition
Locate a range of definitions everyday definitions scholarly definitions
Select one definition Useful Appropriate Relevant Necessary to measure validity
Delineate the dimensions of the concept
Many concepts have a number of different aspects or dimensions should be distinguished
we may use one of these dimensions in the studywe may develop indicators for each dimension
E.g. What are the dimensions of leisure activity?
Conceptual definitions
Importance of definition Determines type(s) of data
Clarifies meaning for researchers, participants and readers“interests” = “likes and dislikes” leisure = “non-utilitarian activities
Concepts and variables
Operationalization finding measurable variables for concepts
Operational definition defines a concept in empirical terms
How can I measure that?
Find Variables for the concept
Definition Concept which varies in type or amount
Gender varies in type Male or female
Time spent engaging in leisure activities varies in amount
Concept which is measurable Concept to which values have been assigned.
Values must be - exhaustive; exclusive
Variables
For most concepts there will be many variables Leisure Activities of interest
Different types of Activities: Engages in boxing Engages in cooking
Amount of time engaging in leisureLocation of activityLevel of organisation
Indicators
Indicators become the focus of questions asked and evidence gathered
An empirical observation that can be taken as evidence of particular attributes of a variable eg male or female is an indicator of gender Marital status? Educational level?
Indicators
To study the compassion of LIS570 students “a feeling of deep sorrow for living things stricken by misfortune” “a strong desire to alleviate the pain of living things”
indicators cry when people die in movies vomit at motor accidents feel sad when the local possum stops coming for food
Indicators
How many indicators should we use? How do we develop indicators?
Measures developed in previous studies observation, unstructured interviews informants
Evaluating indicators
Validity the indicators measure the concept that we think they are
measuring appropriateness and relevance of the indicators
eg is educational level a valid indicator of social status?
reliability we can rely on the answers that people give to the
questions that we ask
Levels of measurement
Any variable is composed of 2 or more categories or attributes eg sex (male/ female); country of birth
(Australia, USA, NZ etc)
Level of measurement refers to how the categories of the variable relate to one another
Levels of measurement
Nominal - measuring a variable at this level involves naming the calibration units (1) = Male Value Label
(2) = Female Value Label
Sample data: 2 2 1 1 2 2 1 1
Levels of measurement
Ordinal - involves arranging the calibration units into a logical order of rank
Age (number of years) (1) 18-24 (2) 25-31 (3) 32-38 (4) 39-45
There is an order in the calibrations without any assumption that the distances between each calibrating unit are equal
Levels of measurement
Interval e.g.Number of hours;
Values designate quantity; 2 more than 1 less than 3; Person 2 has more of the characteristic number of hours exercising (5 hours) than person 1 (2 hours); Difference between them is 3 hours
Has order but also involves specifying an equal distance between each successive unit
Levels of measurement
Ratio specifies a distance between each successive
calibration assumes there is a point at which the variable
literally does not exist eg height and weight
Which level to aim for (De Vaus)
Interval or ratio levels precise averages can be calculated
ie what is the average sex of students at UW? More powerful and sophisticated techniques of analysis
are available higher levels of measurement provide more information interval level measures can be converted to ordinal or
nominal level but not vice versa
Summary - descending the ladder of abstraction