Class Notes (1,100,000)
CA (630,000)
U of G (30,000)
SOAN (900)
SOAN 2120 (400)
Lecture

SOAN 2120 Lecture Notes - Face Validity


Department
Sociology and Anthropology
Course Code
SOAN 2120
Professor
Scott Schau

This preview shows half of the first page. to view the full 2 pages of the document.
SOAN LECTURE September 29
Indexes, Scales and Typologies
-Most common in survey research and other quantitative methods
-Often used in quantitative research because:
oDifficult to devise a single measure of complex concepts
Especially true with regard to attitudes and orientations
Researchers devise several items, each providing some indication of the
variable
-A single indicator provides a valid and reliable measure of some variables
oE.g. Marital status, number of children
-Other variables are more difficult and complex to measure
oSome may require > 1 indicator of a variable
E.g. depression
oSome may involve a # of dimensions that calls for multiple indicators
E.g. compassion
oSome may be concerned with degree to which a variable is present
E.g. physically fit
Indexes and Scales: Differences
-Way scores are assigned in each:
oIndex- accumulate scores assigned to individual attributes
oScale- assign scores to pattern of responses
Uses differences in intensity among the attributes of the same variables
to identify distinct pattern of response
Some items reflect a relatively weak degree of the variable while
other reflect something stronger
Indexes and Scales
-2 misconceptions about scaling:
o1) A particular sample of observations determines whether the combination of
several data items results in a scale
Certain items may form a scale within one sample but not within another
o2) Use od specific scaling techniques does not guarantee the creation of a scale;
instead such techniques help determine whether of not a set of items constitute a
scale
Index Construction
-Main steps
oSelect possible items
oExamine their empirical relationships
oScore your index
oValidate it
Step 1: Select Item
-Select items for a composite index, created to measure some variable
oFace Validity
Each of the items should appear on its face to indicate the concept you
are interested in measuring
oUnidimensionality
A composite measure should represent only one dimension of a concept
oGeneral or Specific
Nature of the items included determine how specific of general the
variable is measured
You're Reading a Preview

Unlock to view full version