Path: Janda: Political Parties, Home Page > Part 1: Table of Contents > Chapter 2
Chapter 2: Variable Coding and Data Quality (pp 12-18), this is p. 16
(you can navigate to other pages by clicking on page numbers below)
p. 12
p. 13
p. 14
p. 15
----
p. 17
p. 18

TABLE 2.1
Adequacy-Confidence Scale
Code
Category Label
Conceptual Definition
Operational Definition
Blank

Nonexistent, uncoded

Party did not exist in time period, or variable was not coded because of lack of funds.

0

Inapplicable

Variable does not apply to the party coded.

1

Inadequate: no data

No information is contained in the file on the variable being considered.

2

Inadequate: disagreement

Disagreements in the file are not resolvable by reference to source credibility. The disagreement might be resolved by more data, but the information in the file is inadequate.

3

Barely adequate: lowest confidence

It is possible to cite this code as the most probable among alternatives, but further research could easily produce a finding at great variance from this one.

Two situations can produce this code: (1) There is a 1:1 division between sources with a "great" discrepancy1 in the suggested codes, but one code could be favored on the basis of source credibility, and (2) data are incomplete in some way, but a code can be inferred from available information.

4

Adequate: low confidence

Disagreement in the literature suggests that the code might not be supported by further research, although the alternative is not greatly discrepant.

There is a 1:1 division between sources with a "medium" discrepancy' in suggested codes, but one code can be favored on the basis of source credibility.

5

Adequate: low to medium

There is no strong agreement in the literature on this particular code, but further research is likely to support the code or one close to it.

Three situations can produce this code: (1) No source has complete information, but a summary code can be made from data from two or more incomplete sources, (2) there is a 2:1 division between sources without regard to degree of discrepancy, and (3) there is a 1:1 division between sources with a "small" discrepancy1 in suggested codes, but one code can be favored on the basis of source credibility.

6

Adequate: medium

The code is not extensively documented in the literature, but there is no disagreement in evidence. Further research would likely support the code, but there are no strong grounds to rule out the possibility of disagreement.

One source cites the summary code with no disagreement in evidence.

7

Adequate: medium to high

Although the code is quite well documented, the judgment is placed in some doubt because it is not unanimous. Disagreements might occur in further research, but the code would likely be supported

There is at least a 3:1 division between sources, without regard to the degree of discrepancy, and the overwhelming evidence favors the code.

8

Adequate: high confidence

Since documentation of the code is good and no disagreements are apparent, it is probably accurate, although additional documentation is desirable.

Two sources agree on the code and no source disagrees.

9

Adequate: highest confidence

The variable code is extremely well documented and no disagreements are apparent; belief in the accuracy of this code is about as high as one could expect in the absence of original field research.

Three or more sources agree on the code and no source disagrees.


1The degree of discrepancy is applicable only to variables of an ordinal or interval nature. Whether a discrepancy is to be classified as "small," "medium," or "great" depends on the particular variable and is established in the operational definitions for each variable, which must be referred to in order to determine or interpret the degree of discrepancy.

go to page 17