Goal setting and archery progress: Barnett 1979 study

In three beginners' classes at Wayne State University in Detroit, students were randomly assigned to two groups for the 10 weeks of the course. Everyone had a weekly 10-minute group meeting with the instructor; in one group, at the end of the meeting, each student wrote down their own technical and score goals on a printed sheet, while in the other the same problems were discussed without setting goals.

On the final test at 20 yards the goal-setting group averaged 40.26 points per six-arrow end, the other group 36.62; at the mid-course test in week six the two groups were level (32.97 and 33.18). Taking the initial test into account, the adjusted mean across the two tests was 37.05 against 31.95, a significant difference (F 5.31, p below 0.05). The group by time interaction, meaning faster improvement in the goal-setting group between weeks six and ten, was not significant (F 3.85).

Each group started with 20 students, but only those with all three tests were analysed: 18 in the goal-setting group and 12 in the other. The course ran two 50-minute classes a week, always with the same instructor, who also led the meetings. The goal sheet listed technical points grouped by phase of the shot (stance, nocking, draw, aim, release, follow-through), made harder in the second half of the course, with space for personal goals; the numerical goal was the increase in score for one six-arrow end during the week. The group without the sheet heard the same points out loud, but was not asked to set goals.

The initial test, in week one after very little instruction, was the mean of two ends at 20 yards (about 18 metres); the mid-course and final tests were the mean of four ends. After the first test the students restarted at 10 yards and the distance was increased gradually over the first five weeks. The analysis is an ANCOVA with the initial test as covariate.

Drop-out is the main problem, and it is badly unbalanced: 2 students lost from the goal-setting group and 8 from the other, 7 of whom had stopped attending and left the course. The authors take this as a sign that goals keep motivation up, but drop-out also changes who is in each group, and with 12 people in the comparison group it takes little to shift a mean. The initial test scores are not reported; the gap between raw means (36.61 against 34.90) and adjusted means (37.05 against 31.95) shows that the group without goals started higher, and much of the advantage comes from the adjustment.

There was a single instructor, who knew which student belonged to which group. The study does not say what target face or scoring system was used, nor what bows the students shot. There is no follow-up after the course ended, and motivation, which the authors put forward as the mechanism, was not measured.

It is one of the earliest studies of goal setting in archery, with complete beginners on a university course. It suggests that writing down a technical goal and a score goal each week, rather than only discussing them with the instructor, goes with slightly higher scores at the end of the course and fewer drop-outs. With 30 students, a non-significant interaction, a comparison group reduced to 12 and an adjustment that weighs heavily on the result, it remains a pointer rather than a measure of the effect. It says nothing about experienced archers or competition, where goals play a different role.