Skip to main content

Table 2 Univariate Differences between training sample and validation sample

From: An application in identifying high-risk populations in alternative tobacco product use utilizing logistic regression and CART: a heuristic comparison

  Training Validation P value
(n = 1584) (n = 792)
Demographics    
Male 657 (27.7%) 337 (14.2%) 0.617
Age (±SD) 42.94 ± 12.39 43.03 ± 12.5 0.880
Race    0.997
  African American 530 (22.3%) 264 (11.1%)
  Latino 524 (22.1%) 262 (11.0%)
  White 530 (22.3%) 266 (11.2%)
Education, % college graduate or higher 550 (23.1%) 288 (12.1%) 0.430
Income, % < $1800/month 614 (25.8%) 329 (13.8%) 0.192
Tobacco Characteristics    
Smoking status (%)    0.263
  Nondaily 799 (33.6%) 402 (16.9%)
  Daily light (1–10 cpd) 373 (15.7%) 205 (8.6%)
  Daily heavy (11+ cpd) 412 (17.3 ) 185 (7.8%)
Menthol smoker 899 (37.8%) 461 (19.4%) 0.500
Cigarettes per day, mean (±SD) 10.03 ± 9.03 9.06 ± 7.69 0.009
Time to first cigarette, % within 30 minutes of waking 900 (37.9%) 449 (18.9%) 0.953
24 hour quit attempts in last 12 months, mean (±SD) 5.54 ± 9.87 6.00 ± 11.93 0.454
Cost    
Price of cigs influenced them to smoke less, % yes 920 (38.7%) 450 (18.9%) 0.557
Price of cigs influenced where they buy cigs, % yes 1100 (46.3%) 566 (23.8%) 0.311
Price of cigs influenced the brand they buy, % yes 685 (28.8%) 360 (15.2%) 0.306
Buy versus borrow cigs, % buy all cigs they smoke 1004 (42.3%) 503 (21.2%) 0.952
Harm Reduction    
Trying to cut down on cigs smoke, % yes 1119 (47.1%) 561 (23.6%) 0.924
Limit cpd to decrease health risk, % yes 730 (30.7%) 371 (15.6%) 0.727
Limit smoking in last year to decrease health risks, % always or often 476 (20.0%) 240 (10.1%) 0.899
Psychosocial    
Depression score, mean (±SD)a 1.99 ± 1.86 1.96 ± 1.82 0.683
Alcohol score, mean (±SD)b 4.02 ± 3.16 3.93 ± 3.03 0.494
Discrimination score, mean (±SD)c 7.03 ± 6.30 7.23 ± 6.44 0.460
  1. aScores range from 0–6 with scores of 3 or higher indicating possible depressive symptoms.
  2. bScores range from 0–12 with scores of ≥4 for men and ≥3 for women indicating possible alcohol misuse.
  3. cScores range from 0–25 with higher scores indicating greater frequency of discrimination in daily life.