Skip to main content

Table 3 Reliability tests comparing dietary choline intake between quantitative food frequency questionnaire interviews

From: Development and validation of a quantitative choline food frequency questionnaire for use with drinking and non-drinking pregnant women in Cape Town, South Africa

 

Group-level

Individual-level

Summary

Agreement

Agreement

Agreement

Bias

Strength of association

Agreement including chance

Agreement excluding chance

Wilcoxon signed rank test p

Percentage difference

Bland-Altman LOA index

Bland-Altman Spearman correlation p

Spearman correlation2

Tertile classification

Kappa statistic

Group-level tests at “Good” or “Acceptable” (of 4 tests)

Individual-level tests at “Good” or “Acceptable” (of 5 tests)

Same tertile

Opposite tertile

Weighted for tertiles

Above vs. below AI

Heavy drinkers:

 Visit 1 vs. 2 (n = 112)

.001

M -0.4%

4.5%

< .001

.52

54.5%

9.8%

.44

< .001

  

 Interpretation1

Poor

Good

Good

Poor

Good

Good

Good

Acceptable

Poor

2

4

 Visit 1 vs. 3 (n = 78)

.026

M -2.8%

7.7%

< .001

.43

47.4%

14.1%

.28

< .001

  

 Interpretation1

Poor

Good

Poor

Poor

Acceptable

Poor

Poor

Acceptable

Poor

1

2

 Visit 2 vs. 3 (n = 78)

.606

M -18.4%

5.1%

< .001

.56

50.0%

9.0%

.41

< .001

  

 Interpretation1

Good

Acceptable

Good

Poor

Good

Good

Good

Acceptable

Poor

2

4

Abstainers/light-drinkers:

 Visit 1 vs. 2 (n = 72)

.522

M -7.1%

4.2%

.001

.42

56.9%

2.8%

.50

.001

  

 Interpretation1

Good

Good

Good

Poor

Acceptable

Good

Good

Acceptable

Poor

3

4

 Visit 1 vs. 3 (n = 42)

< .001

M 9.1%

4.8%

< .001

.37

47.6%

2.4%

.49

< .001

  

 Interpretation1

Poor

Good

Good

Poor

Acceptable

Poor

Good

Acceptable

Poor

2

3

 Visit 2 vs. 3 (n = 42)

.014

M 16.2%

2.4%

< .001

.50

50.0%

4.8%

.47

< .001

  

 Interpretation1

Poor

Acceptable

Good

Poor

Good

Good

Good

Acceptable

Poor

2

4

  1. 1Interpretation criteria [37]:
  2. Wilcoxon signed rank test: Good: p > 0.05; Poor: p ≤ 0.05
  3. Percentage difference: Good: 0.0 – < 11%; Acceptable: 11 - 20.0%; Poor: > 20%
  4. Bland-Altman LOA index (% outside of limit of agreement): Good: ≤ 5%; Poor: > 5%
  5. Bland-Altman Spearman correlation coefficient: Good: p > 0.05; Poor: p ≤ 0.05
  6. Spearman correlation coefficient: Good: ≥ 0.50; Acceptable: 0.20 – 0.49; Poor < 0.20
  7. Tertile classification (% in same tertile): Good: ≥ 50%; Poor: < 50%
  8. Tertile classification (% in opposite tertile): Good: ≤ 10%, Poor: > 10%
  9. Weighted Kappa statistic: Good: ≥ 0.60; Acceptable: 0.20 – 0.59; Poor: < 0.20
  10. 2p < .001 for all Spearman values
Back to article page

Nutrition Journal

ISSN: 1475-2891

Contact us