Social Referencing Facilitates Fitness with Fitbit: Competition Wins, Collaboration Places

Lewis & Clark College HCI LAB - WPA / 05 - 2015

We conducted an experiment exploring the impact that social referencing, using an electronic activity tracker, has on persistence and success. Specifically, this study asks whether social referencing utilizing competition or collaboration is more effective at increasing step count and activity level. Sixteen young adults tracked their step count for a three week period using an electronic activity tracker (Fitbit Zip). For the first week, we asked participants to wear the Fitbit Zip all day and check in on the company website each evening to see how many steps they had taken that day. On the basis of their first week’s activity, we created four teams of four people with two teams composed of people with a relatively high step count of 10,000 per day (High Steppers) and two teams (Low Steppers) who only averaged 6,000 steps per day. During weeks two and three they competed in two social referencing conditions (in a counterbalanced order) with their four-person team. They used either competition (for small prizes won by 1st and 2nd place finishers) or collaboration (donating to a local charity based on the group total step count for the week). A 2 (Activity Level) x 3 (Condition) Mixed Model ANOVA was used for the number of steps walked and the active minutes per day in each phase of the experiment. Both social referencing (cooperation and competition) conditions significantly increased the number of steps taken per day compared to the baseline condition (p <.01). Post hoc tests showed that competition (M = 11,018) was slightly more effective than collaboration (M = 9735) (p =.08). Both social facilitation conditions also increased the number of active minutes obtained daily (p <.01). A paired sample t-test revealed that participants in the Low Steppers group had significantly more active minutes in the competitive condition (M=372) than in the cooperative condition(M=248) (p=.026). No significant interactions were found for either number of steps walked or active minutes per day. In future studies we plan to look into the effects of different types of competition and cooperation to optimize the effectiveness of social referencing.


Decades of research tell us that manipulating physical objects is essential to early childhood development. There is a growing use of digital technology for early childhood education. This study evaluates the use of an iPad app called Tiggly Draw. It combines the use of physical shapes and digital technology to teach preschoolers basic shape recognition and to support creative play. The game play of the app involves children drawing with canonical shapes for (circles, squares, triangles, and stars) and a selection of features that can be dragged onto the shapes to create different designs. We had 50 three to four year olds play, with one of two versions of Tiggly Draw, for eight minutes. The control condition (app only) consisted exclusively of screen interactions with children dragging the shapes and features from on screen icons. The experimental condition (tiggly shapes) used 4 soft plastic shapes that were physically pressed on the screen to create the shapes. In both conditions, upon creating a shape the app writes and speaks the name of the shape. Sounds and animations also accompany the placement of features on the drawings. Our hypotheses were that both shape recognition and creativity would be aided by the combination of digital and physical interaction. Shape recognition was measured through a pretest/post timed trial that involved placing the 4 shapes into identical shaped holes in a plastic toy. Using a 2 x 2 ANOVA we found a significant interaction. Three year olds showed significantly greater improvement using the tiggly shapes version while four year olds showed no difference (p= .02). An analysis of the game play found that the younger children also created more shapes than the older children (p= .01). Our primary measure of divergent creativity was the number of unique features that children used in their 8 minutes of play. Here, we found a distinct advantage for the four year olds with both main effects for condition (p = .01) and an interaction (p = .001). The use of tangible tiggly shapes produced more divergent play than the app only version for four years olds.