Scoring the All-Day Screener

On this page...

• How to Score the Screener
• Earlier Changes in Scoring
• How to Handle Missing Data

How to Score the Screener

For the All-Day screener, scoring involves a series of operations that are shown below and implemented in the All-Day Screener Pyramid Servings SAS Program and the All-Day Screener MyPyramid Cup Equivalents SAS Program.

1. Express each reported frequency as a daily average. To do this, standardize the midpoint of each frequency category to the number of times per day.

   Frequency Response	Times Per Day
   Never	0.0
   1-3 times per month	0.067
   1-2 times per week	0.214
   3-4 times per week	0.5
   5-6 times per week	0.786
   1 time per day	1.0
   2 times per day	2.0
   3 time per day	3.0
   4 times per day	4.0
   5 or more times per day	5.0

   
   

2. Assign fruit and vegetable Pyramid/MyPyramid servings to each portion size category.

   1. Using 2005 MyPyramid definitions of fruit and vegetable cup equivalents:

      Food	MyPyramid Cup Equivalents for each Portion Size Category
      	1	2	3	4
      Juice	.5	1.0	1.625	2.5
      Fruit (units)	.25	.5	1.0	1.5
      Fruit (cups)	.25	.5	1.0	1.5
      Lettuce salad	.25	.5	1.0	1.5
      French fries	.2	.5	.75	1.3
      Other white potatoes	.25	.75	1.2	2.0
      Dried beans	.25	.75	1.25	2.0
      Other vegetables	.25	.75	1.5	2.25
      Tomato sauce	.25	.5	1.0	1.5
      Vegetable soups	.3	1.0	1.6	2.25

      
      

   2. Using 1992-2004 definition of Pyramid servings:

      Food	Pyramid Servings for each Portion Size Category
      	1	2	3	4
      Juice	.75	1.33	2.17	3.33
      Fruit (units)	.75	1.0	2.0	2.5
      Fruit (cups)	.75	1.0	2.0	2.5
      Lettuce salad	.5	1.0	2.0	3.0
      French fries	1.25	2.3	3.1	4.8
      Other white potatoes	.8	1.5	2.4	3.5
      Dried beans	.75	1.5	2.5	3.5
      Other vegetables	.75	1.5	3.0	4.5
      Tomato sauce	.36	.72	1.45	1.7
      Vegetable soups	.75	1.36	2.27	3.2

      
      
3. Compute the average daily fruit and vegetable Pyramid/MyPyramid servings for each food group by multiplying the average daily frequency (from #1) by the number of Pyramid/MyPyramid servings for the portion size (as determined in #2). To estimate the total daily number of Pyramid/MyPyramid servings, sum across all food groups.
4. Note: Question 10 -- "How often over the past month did you eat mixtures that included vegetables?" -- is not included in this algorithm. This question may be helpful in identifying individuals who eat a lot of their vegetables in mixtures. If amounts of vegetables eaten in mixtures are harder to report, their reports may be somewhat less accurate. However, this supposition has not yet been tested.

Earlier Changes in Scoring

On July 12, 2001, the following changes were made to the calculations used to estimate fruits and vegetables from the screener data:

1. In the calculations to convert the frequency categories to number of times per day:
   • 1-3 times per month was changed from .066 to .067 times per day,
   • 3-4 times per week was changed from .499 to .5 times per day, and
   • 5-6 times per week was changed from .784 to .786 times per day.
2. In the assignment of fruit and vegetable Pyramid servings to portion size categories:
   • the highest portion size category for juice was changed from 2.5 to 3.33 Pyramid servings and
   • the highest portion size category for lettuce was changed from 2.5 to 3.0 Pyramid servings

If you need the previous version of the SAS program, please send an e-mail to rfab@mail.nih.gov.

How to Handle Missing Data

1. Assigning reasonable data to individual questions with missing answers:

   1. Frequency of intake for individual foods
      Assume Never=0 because many people do not fill in any bubble when they do not consume a food.

      If portion size for that item is answered and the frequency question is not answered, continue to treat as Never.

   2. Portion size
      Note: portion size is not missing if the frequency question stem above it is answered Never or assigned Never.

      If a vegetable, assign that individual's most frequently reported portion size category for all other separate vegetable questions. [Exclude "other vegetables" from this calculus because the portions reported there may reflect portions of all other vegetables rather than a single vegetable at a time.] In order to find the most frequently reported portion size category, weight each portion size category by the relative frequency with which each portion is reported, so that the portion sizes associated with more frequently consumed vegetables are given more weight than the portion sizes associated with less frequently consumed vegetables.

      If a fruit, assign the portion size category reported for the other fruit question. If neither fruit juice or fruit portion sizes are reported, assign the most frequently reported portion size for vegetables.

2. Criteria for excluding cases (Note that these exclusion criteria are being developed, and may vary from study to study):
   • If frequency information on fruit or other vegetables is missing.
   • If nearly all questions are unanswered.
   • If more than one response to a question is provided.