Data Dictionary Example

Example Table

subject_id Subject's ID Eyedisease.doc integer
case_control Is paricipant case or control? Eyedisease.doc
age Age of disease onset Eyedisease.doc integer years 25 90 90=">= 90"
height Height of subject Eyedisease.doc integer inches 59 75
sex Gender of participant Main_exam.txt enumerated integer 1 = Male; 2 = Female; NA = Not applicable; UNK = Unknown
weight Weight of participant Diagnosis.pdf decimal pounds 99.5 500.8 -1 = missing; -2 = not assessed 1
mothocc Occupation of mother Background.pdf string
fatherocc Occupation of father Background.pdf string
gluc_conc Concentration of blood glucose Eyedisease.doc decimal mg/L 700.45 1500.08 2
dementia Dementia status Psych.doc string Y = yes; ? = maybe; N = no;. = missing Dementia was defined using DSM-IV criteria. Symptoms be present for >= 6 months.

Explanation of Column Headers

Column Header Description
VARNAME Variable name
VARDESC Variable description
DOCFILE Name of document associated with the variable
TYPE Integer (1,2,3,4...), enumerated integer (integers are coded for non-integer meaning, see CODE_MEANINGS), decimal (0.5,2.5,...), string (reporter, doctor, lawyer...)
UNITS Unit of variable if applicable, otherwise leave blank
MIN Logical minimum value of variable if applicable
MAX Logical maximum value of variable if applicable
CODE_MEANING Description of enumerated integer, use equal sign (=) between code and meaning, and use semicolons (;) only to delimit code
RESOLUTION (Optional) Measurement resolution - the number of decimal places to which a measured value is presented in the data
COMMENT (Optional) Additional information not included in the VARDESC that will further delineate the variable. If additional comments are needed beyond COMMENT3, create columns (COMMENT4, COMMENT5, etc).

VARNAME and VARDESC are required columns for all data dictionaries; all other columns should be included when applicable