Applied Human Conduct Review — Issue No. 0010
This issue looks at the game elements AI coaches add to practice — streaks, points, levels, badges — and asks what each one rewards and what happens to the people it does not suit. Four brief reports follow adults learning a language, walking with neighbors, learning spreadsheet analysis at work, and taking up an instrument. In each, the better design was not the one with the most engagement at the start. It forgave a missed day instead of resetting the count, counted a group's progress instead of each member's, awarded points for skill shown instead of time spent, or withdrew its rewards early and gradually instead of letting them run until the program ended. Each of those choices cost something early on, usually in the engagement figures programs report, and someone accepted that cost: program operators, the most active walkers, or learners who found a course harder for a while. As with all issues of Applied Human Conduct Review, reports favor concrete, replicable behavioral comparisons over normative argument.
Papers in This Issue
We randomly assigned 402 adults starting a 12-week AI-coached language course to one of three streak designs and followed their logged practice to week 24. In "strict-streak" accounts (n=134), the coach counted consecutive practice days and reset the count to zero after a missed day. In "repairable-streak" accounts (n=136), one missed day a week was absorbed, and a further missed day could be repaired with a five-minute catch-up session within two days. "No-streak" accounts (n=132) used the same coach and lessons with no counter. Strict-streak participants practiced most in the first four weeks (5.6 sessions a week, against 5.2 for repairable-streak and 4.4 for no-streak participants), but over the full 12 weeks repairable-streak participants completed a mean 46.9 sessions, against 44.6 for strict-streak and 39.8 for no-streak participants. By week 12, most participants in every group had a break of two or more days; 59% of repairable-streak participants who had one returned within seven days, against 46% of strict-streak participants (a 28% gap) and 52% of no-streak participants. In weeks 21 to 24, 40% of repairable-streak participants were still practicing every week, against 31% of strict-streak participants (a 27% gap) and 33% of no-streak participants (a 19% gap). Participants who had missed a day by week 6 rated how badly they felt about it at 4.1/7 under the strict streak, 3.2/7 under the repairable streak (strict-streak ratings 28% higher), and 2.9/7 with no streak.
In a 16-week community walking program, 48 groups of five or six members (276 people) used the same AI coach and were randomly assigned by group to one of three game layers. In "shared-quest" groups (16 groups, 93 members), the coach set a weekly distance target for the group as a whole, showed only the group's combined progress, and suggested that members walking ahead of their usual pace invite another member along. In "individual-points" groups (16 groups, 92 members), each member earned points for their own distance, visible only to themselves. "No-game" groups (16 groups, 91 members) received the coach's walking suggestions alone. In weeks 13 to 16, 62% of shared-quest members still logged at least three walks of 20 minutes or more every week, against 51% of individual-points members (a 22% gap) and 49% of no-game members (a 26% gap). Among members in the least active third of their design at baseline, the figures were 48%, 39% (a 25% gap), and 40% (a 21% gap). Consenting shared-quest members logged 1.9 joint walks with another group member a month, against 1.5 for individual-points members and 1.6 for no-game members. Members in the most active third of the shared-quest design walked 6% less than their counterparts in individual-points groups, and 8 of 31 reported at least once feeling they were carrying the group.
In a 10-week workplace course on spreadsheet analysis taught by an AI tutor, 348 employees from seven organizations were randomly assigned to one of three reward designs. In "mastery-gated" accounts (n=117), points and new levels came only from passing short applied checks set by the tutor, with unlimited retries and a hint after each failed attempt. In "time-points" accounts (n=116), points came from minutes practiced and lessons completed, and levels from accumulated points. "Plain" accounts (n=115) had the same lessons with no points or levels. Time-points learners logged the most practice, a mean 142 minutes a week, against 118 for mastery-gated and 112 for plain learners, and 71% of them finished the course within 10 weeks, against 58% of mastery-gated learners and 61% of plain learners. Three months after the course ended, 331 learners took an applied task scored by two raters blind to design; mastery-gated learners scored 64.2/100, against 52.8 for time-points learners (a 22% gap) and 55.1 for plain learners (a 17% gap). Mastery-gated learners rated the course as more frustrating, 3.6/7 against 2.9/7 for time-points learners (a 24% gap) and 3.1/7 for plain learners (a 16% gap). Mastery-gated learners who did not finish within 10 weeks still scored 60.1, above the time-points average.
We randomly assigned 294 adult beginners in a 16-week AI-coached keyboard and guitar program to one of three game-layer designs and followed their logged practice to week 28, twelve weeks after the program's game features closed for everyone. In "fading" accounts (n=99), points, a weekly target meter, and milestone badges ran for six weeks and were then removed one at a time over weeks 7 to 10, while the coach began asking each learner weekly what they wanted to work on next. Fading learners who went a week without practice after the meter's removal in week 9 were offered it back for two weeks. In "persistent" accounts (n=98), the full game layer ran until week 16, and "no-game" accounts (n=97) had none. In weeks 7 to 10, persistent learners practiced 3.9 sessions a week, against 3.4 for fading learners and 3.2 for no-game learners. When the game layer ended at week 16, persistent learners' practice fell from 3.8 to 2.8 sessions a week, a 26% drop, while fading learners' fell from 3.3 to 3.0. In weeks 25 to 28, 45% of fading learners still practiced at least twice every week, against 36% of persistent learners (a 27% gap) and 38% of no-game learners (a 19% gap). At week 16, fading learners rated their interest in playing for its own sake 5.1/7, against 4.3/7 for persistent learners (a 19% gap) and 4.6/7 for no-game learners. Of the 17 fading learners who took the target meter back, 7 were still practicing at least twice a week in weeks 25 to 28.