Abstract:

According to Agenda-Based Regulation Model (ABR), individuals will comprehensively analyze various factors such as task objectives, task constraints to construct the learning agenda, which is used to prioritize the study items and the amount of time needed to study. However, the main concern of the previous studies is the value presented as a reward outcome (reward obtained after successfully memory), leading to a lack of valid examination of whether reward expectation (prediction of reward outcome) affects the agenda construction and memory performance. The present study was to supplement the reward expectation into the ABR model by verifying whether a sufficiently high reward expectation can replace difficulty to become a dominant influence on JOLs and study time allocation when forming an agenda.

Experiment 1

60 participants joined this experiment. Added a control group on the basis of Soderstrom and McCabe's (2011), 2 (difficulty: easy, hard) × 2 (reward expectation: yes, no) and 2 (difficulty: easy, hard) × 2 (reward outcome: high, low) experiments were designed to examine the effect of reward expectation and difficulty on JOLs and memory rates. Participants have to give JOLs to the word pairs that they studied under a time limited condition, and finish a test in the end.

The results showed that reward outcomes facilitated the memory performance (F(2, 46) = 9.25, p < 0.001, η²_p = 0.29) and JOLs (F(2, 46) = 5.18, p = 0.009, η²_p = 0.19) of both easy and hard word pairs (see Figure 1), and reward expectation only improved the memory performance of easy word pairs (F(1, 53) = 4.51, p = 0.038, η²_p = 0.08) without significant effects on JOLs (F(1, 53) = 1.70, p = 0.198) (see Figure 2).

Experiment 2

60 participants in this experiment. Only shifted limited-time learning to self-paced learning to examine the effects of reward expectation and difficulty on the study time allocation.

The results showed that reward outcome affected the JOLs (F(2, 46) = 5.18, p = 0.009, η²_p = 0.18) rather than memory performance (F(2, 46) = 0.01, p = 0.986) (see Figure 3). But reward expectation promoted both JOLs (F(1, 50) = 4.90, p = 0.031, η_p² = 0.09) and study time allocation(F(1, 51) = 4.76, p = 0.034, η²_p = 0.09), thus improving the memory performance (F(1, 50) = 6.51, p = 0.014, η²_p = 0.12) (see Figure 4). JOLs and study time allocation of hard word pairs in condition with reward expectation are higher than with no reward (see Figure 5).

Experiment 3

18 participants in this experiment. Two sequences of value outcomes (1, 6, 12 or 1, 3, 6) enabled participants to generate high gradient and low gradient reward expectations respectively to further investigate the impact of the magnitude of reward expectation.

In self-paced learning in Experiment 3, the influence of difficulty on study time not significant any more (F(1, 14) = 3.87, p = 0.069), reward expectation beyond difficulty becomes the main factor affecting the study time allocation (F(1, 14) = 4.55, p = 0.050, η²_p = 0.25) (see Figure 6).

The above results proved that reward expectation is a contributing factor in ABR model. Individuals synthesize reward expectation, reward outcome and difficulty while constructing a learning agenda, and reward expectation overrides difficulty as the dominant factor in agenda construction when it is sufficiently large. However, the effects of reward expectation and reward outcome on memory performance, study time allocation, and JOLs were modulated by the learning conditions.

Key words: reward expectation, judgments of learning, study time allocation, metamemory, memory