Background: In many studies, it is of interest to identify population subgroups that are relatively homogeneous with respect to an outcome. The nature of these subgroups can provide insight into effect mechanisms and suggest targets for tailored interventions. However, identifying relevant subgroups can be challenging with standard statistical methods. Main text: We review the literature on decision trees, a family of techniques for partitioning the population, on the basis of covariates, into distinct subgroups who share similar values of an outcome variable. We compare two decision tree methods, the popular Classification and Regression tree (CART) technique and the newer Conditional Inference tree (CTree) technique, assessing their performance in a simulation study and using data from the Box Lunch Study, a randomized controlled trial of a portion size intervention. Both CART and CTree identify homogeneous population subgroups and offer improved prediction accuracy relative to regression-based approaches when subgroups are truly present in the data. An important distinction between CART and CTree is that the latter uses a formal statistical hypothesis testing framework in building decision trees, which simplifies the process of identifying and interpreting the final tree model. We also introduce a novel way to visualize the subgroups defined by decision trees. Our novel graphical visualization provides a more scientifically meaningful characterization of the subgroups identified by decision trees. Conclusions: Decision trees are a useful tool for identifying homogeneous subgroups defined by combinations of individual characteristics. While all decision tree techniques generate subgroups, we advocate the use of the newer CTree technique due to its simplicity and ease of interpretation.
1 January 2010 in International Journal of Behavioral Nutrition and Physical Activity
Objective: To evaluate the effects of lowering prices and increasing availability on sales of healthy foods and beverages from 33 vending machines in 4 bus garages as part of a multicomponent worksite obesity prevention intervention. Methods: Availability of healthy items was increased to 50% and prices were lowered at least 10% in the vending machines in two metropolitan bus garages for an 18-month period. Two control garages offered vending choices at usual availability and prices. Sales data were collected monthly from each of the vending machines at the four garages. Results: Increases in availability to 50% and price reductions of an average of 31% resulted in 10% to 42% higher sales of the healthy items. Employees were mostly price responsive for snack purchases. Conclusions: Greater availability and lower prices on targeted food and beverage items from vending machines was associated with greater purchases of these items over an 18-month period. Efforts to promote healthful food purchases in worksite settings should incorporate these two strategies.
This study investigated the acute effects of repeated walking sessions within green and suburban environments on participants’ psychological (anxiety and mood) and cognitive (directed-attention) outcomes. Twenty-three middle-aged adults (19 female) participated in a non-randomized crossover study comprised of once-weekly 50-min moderate-intensity walking sessions. Participants walked for three weeks in each of two treatment conditions: green and suburban, separated by a two-week washout period. Eleven participants completed green walking first and 12 suburban walking first. For each walk, we used validated psychological questionnaires to measure pre- and post-walk scores for: (1) mood, evaluated via the Positive and Negative Affect Schedule (PANAS); (2) anxiety, assessed by the State-Trait Anxiety Inventory (STAI-S); and (3) directed-attention, measured using the visual Backwards Digit Span test (BDS). Repeated measures linear mixed models assessed pre- to post-walk changes within-treatment conditions and post-walk contrasts between-treatment conditions. Results indicated that anxiety decreased after green walking and increased after suburban walking (−1.8 vs. +1.1 units, respectively; p = 0.001). For mood, positive affect improved after green walking and decreased after suburban walking (+2.3 vs. −0.3 units, respectively; p = 0.004), and negative affect decreased after green walking and remained similar after suburban walking (−0.5 vs. 0 units, respectively; p = 0.06). Directed-attention did not improve from pre- to post-walk for either condition. Our results suggested that green walking may be more effective at reducing state anxiety and increasing positive affect compared to suburban walking.