Learning a Task and Recalling Its Practice

Keep the two records in their own time

Scoville and Milner’s 1957 report distinguishes impaired recent memory from other preserved abilities. Its conclusion also notes that the hippocampal region was not removed alone, leaving the amygdala’s separate role unresolved.

The later Rockefeller account describes improvement on mirror tracing without recollection of the practice sessions.

These are related records, not identical measurements. The later task account should not be silently assigned to the earlier paper or used to claim that its surgery isolated one structure’s contribution.

Sources for this section: [1] · [2]

Define two outcomes in an invented log

Imagine one fictional participant completing three practice blocks of 20 attempts each at fixed task difficulty. Each attempt is scored once as error or no error. The recorded error counts are 12, 8 and 5. After each block, a separate recognition check has 10 items about the practice session, with zero correct responses each time. Assume these are the supplied scores; do not identify the person with H.M. or any real patient.

An error count measures task performance under the specified protocol. The recognition check measures answers to its own questions. Neither is automatically a measure of every kind of memory. The three blocks are repeated observations of one fictional person, not three independent people.

Sources for this section: [1] · [2]

Calculate change and keep the denominator

The task error proportions are 12/20=60%,8/20=40% and 5/20=25%. From the first to last block, the count falls by 7 errors. The error proportion falls by 35 percentage points. Relative to the first count, the reduction is 7/12, approximately 58.3%. These are compatible descriptions of the same supplied change, using different bases.

The session-recognition results remain 0/10 in each block. Adding them gives 0 correct answers across 30 supplied items, but the item total does not turn one person into a population sample or validate the questions. Do not combine the two task scales into a single percentage without an explicit scoring rule.

Sources for this section: [2]

Exercise and reference answer

An invented caption says: “Memory improved by 58.3%, proving the participant remembered all sessions and locating the change in one brain structure.” Rewrite it using the supplied evidence and identify what extra information its last two claims would require.

Reference answer: errors fell from 12 to 5 across equal 20 attempt blocks, a 58.3% relative reduction in errors; session-recognition scores were 0/10 each time. The log supports an improvement on the specified performance measure, not remembered practice. It supplies no brain measurements or intervention that isolates one structure. The caption changes both the outcome and the level of explanation without evidence.

A fuller investigation would need appropriate comparison conditions, task validity and a design suited to the proposed inference. This is a logical and arithmetic exercise, not a diagnostic test, clinical recommendation or reanalysis of a historical patient’s raw results.

Sources for this section: [1] · [2]