**Statistical Analysis:** 1. The same logistic regression analysis as used in Gluth et al. (2015) will be conducted to assess the degree of each participant’s memory bias. That is, we will test to what extend a subject tends to prefer remembered to forgotten items by a logistic regression analysis that is restricted to those decision trials, in which one option was remembered and one option was forgotten. The dependent variable of this logistic regression analysis is the decision (i.e., whether the remembered = 1 or the forgotten = 0 option was chosen). The independent variable of this logistic regression analysis is the standardized subjective value of the remembered option. The logistic regression will be conducted separately for each participant, and subject-specific intercepts and slopes will be derived. The memory bias is given by the estimated intercept of this logistic regression, which we predict to be larger than 0 across participants. A one-sided one-sample t-test against 0 will be used to test for significance of the memory bias (cf. Gluth et al., 2015). 2. A linear regression analysis will be conducted to test for the dependency of participant’s subjective belief on value. The independent variable is the standardized subjective value of each option. The dependent variable is the standardized estimated belief in memory performance of each option. The linear regression analysis will be conducted separately for each participant, and subject-specific intercepts and slopes will be derived (though intercepts are expected to be = 0 given that both the dependent and independent variable are standardized). The dependency of belief in memory performance on the subjective value is given by the slope parameter, which we predict to be larger than 0. A one-sided one-sample t-test against 0 will be used to test for significance of this dependency. 3. A correlation analysis will be conducted to assess the relationship between the two above mentioned regression coefficients across subjects: 1. The acquired regression coefficients of the memory bias 2. The regression coefficients representing individual’s subjective belief. A one-sided Pearson’s or Spearman’s rank correlation test (depending on whether both coefficients are normally distributed) will be conducted to test whether the correlation coefficient is larger than 0. **Transformation:** We will transform subjects’ value ratings as well as the estimated number of correct memories. The rationale of the z-transformation/standardization of subjective preference ratings is that it can straightforwardly illustrate the memory bias (cf. Gluth et al., 2015). The z-transformation of estimated beliefs (but also of subjective value ratings) has the advantage that different usages of the scales (e.g., using the entire range vs. using only a narrow sub-range) do not impact on the estimation of the regression coefficients. **Exclusion criteria:** 1. Demographic exclusion criteria (these criteria will be mentioned on the online recruitment system): a. Age over 39 or below 18 will be excluded b. Participants suffering a mental disorder will be excluded c. Colour blind participants will be excluded d. Participants, which did not follow the pre-experiment requirement not to eat 4 hours previous to the experiment, will be excluded. 2. Value ratings that were extreme (0 or 10) in more than 40% of total cases during the rating task will be excluded. 3. Participants that did not press a button for the sweet. vs. salty evaluation on 40% of encoding task or more will be excluded. Participants with N-Back task performance lower than 70% of total cases, during the task, will be excluded 4. Participants that were too fast (i.e., RT < 250 ms) during the decision task in 30% of the cases or more will be excluded. 5. Participants with recognition performance lower than 30% or higher than 85% of total cases, during the recognition task, will be excluded. 6. Participants with extreme subjective belief values (0 or 6) in 50% of the cases or more, during the subjective belief task, will be excluded. 7. Participants who have at least one estimated regression coefficient (for the memory bias or the belief dependency) which is more than 3 SD lower or higher than the group mean will be excluded. The rationale of the exclusion criteria 2–6 refers to potential failures to accurately estimate the regression coefficients of interest (because a participant did not adhere to the task instructions, criteria 3 and 4; because there are not enough trials; criteria 5 and 6; or because the estimation could be distorted, criteria 2 and 7). The rationale of exclusion criteria 7 is that correlations can be very susceptible to outliers. **Data Collection:** Data will be collected until a total of 67 participants that pass all the exclusion criteria listed above have been tested. The sample size of 67 participants was determined by a power analysis conducted in R using the pwr library (r = 0.3, significant level = 0.05, power level = 0.80, one-sided). The data will be collected at the Department of Psychology, University of Basel. Tehilla Mechera-Ostrovsky will be in charge of scheduling and running the experimental sessions. The data collection will start on March 13th 2017, when the study will be announced on the online recruitment system of the University of Basel. Participants will receive course credits (required for B.Sc. in Psychology) as well as a food snack (depending on the performance and decisions in the experiment) for their participation. **Stimuli:** We will use 48 different food snacks, which are available at well-known Swiss supermarkets, to create choice sets. These snacks will be presented to the participants prior to the experiment. Picture of each snack will be used during the actual procedure. One of these snacks will be given to participants, depending on their performance, following the experiment. **Missing data:** Missing data can occur during the decision phase of the memory-and-choice task, when participants do not adhere to the time limit of making a decision. These trials will be deleted and not analysed. Other missing data that are due to a premature termination of the experiment will lead to the entire exclusion of the participant’s dataset.