Consistent with the position that present findings represent the genuine acquisition of an interpretive bias, a significant four-way interaction was observed between scenario valence condition, scenario ambiguity condition, type of recognition memory statement, and valence of the recognition memory statement, F 1, 74) = 5.09, p = .027, ηp2 = .064.064
To assess whether participants acquired an interpretive bias in line with their assigned valence condition, mean similarity ratings of the positively disambiguated recognition memory statements and the negatively disambiguated recognition memory statements for the ten ambiguous test scenarios were examined.
This comprised a 2 × 2 × 2 × 2 ANOVA with scenario valence condition (positive vs. negative) and scenario ambiguity condition (ambiguity present vs. ambiguity absent) as the between subjects factors, and recognition memory statement valence (positive vs. negative) and recognition memory statement type (disambiguated statement vs. foil statement) as the within subject factors.
Alternatively, the Ambiguity Resolution Account suggests that the presence of ambiguity is critical to modifying interpretive bias and therefore, a three-way interaction will be observed such that the two-way interaction involving scenario valence and recognition memory statement described would only be evident in the ambiguity present condition and not in the ambiguity absent condition.
However, if it were a true interpretive bias, then, we would expect a significant four-way interaction, comprising a significant three-way interaction for the disambiguated recognition memory statements only, and no three-way interaction involving foil statements.
Consistent with the Ambiguity Resolution Account a significant three-way interaction was indeed observed between scenario valence condition, scenario ambiguity condition, and recognition memory statement valence F 1, 74) = 8.79, p = .004, ηp2 = .11.11
These were subjected to a 2 × 2 × 2 mixed model ANOVA with scenario valence condition (positive vs. negative) and scenario ambiguity condition (ambiguity present vs. ambiguity absent) as between-subject factors, and recognition memory statement valence (positive vs. negative) as the within subject factor.
If the resolution of ambiguity is not necessary to achieve change in biased interpretation, as suggested by the Emotional Valence Account, then we would expect a significant two-way interaction between scenario valence and recognition memory statement valence, not further modified by scenario ambiguity condition.
This would be demonstrated in a significant three-way interaction reflecting the same pattern of findings described above, that was not modified further by the type of recognition memory statement (disambiguated statement vs. foil statement).
Examination of the component three-way interactions revealed the significant interaction for the disambiguated recognition memory statements described above F 1, 74) = 8.79, p < .01, ηp2 = .11 and no three-way interaction for the foil statements, F < 1.
This is why Reid calls the evidence of memory immediate: first-personal statements about one's own past are memory statements, not statements made on the basis of memory.
