Call-Center Trainee Recall Audit
You are a training auditor at an insurance call center reviewing a new hire's first week. The trainee handled a 2-minute inbound call while typing notes into a customer-relationship system. The recording shows the caller, Mrs. Delgado, gave her policy number as 4471-828 in two groups, requested a callback after 2:30, and reported water damage from a burst pipe. The floor is open-plan, measured at 62 dBA, and the caller spoke with a regional accent. Immediately after the call the trainee recalled the policy number as 4471-882, the name correctly, the callback as "after 3," and the reason correctly. On a 10-item post-call quiz the trainee scored 4/10. The trainer concludes the trainee has "a bad memory" and proposes extra memorization drills. You must decide what the error pattern actually supports and what to change in the audit before making a training recommendation.
1. Which recalled details are most diagnostic of an auditory memory problem, and which point instead to reconstruction or divided attention? Justify your classification.
Read the worked answer
The pattern matters more than the total score. Name and reason correct shows the caller's identity and gist were encoded and retained; a global auditory memory failure would not spare both. Policy number 4471-828 to 4471-882 is a within-string order error: the digits are preserved but sequence position is lost. That is consistent with limited-capacity phonological storage under divided attention from typing, not with a semantic memory problem. "After 2:30" to "after 3" is a schema-driven rounding error, a reconstruction rather than a stored verbatim trace. So the evidence supports an encoding and phonological-store bottleneck plus retrieval reconstruction, not "bad memory" in general. The 4/10 quiz score is uninterpretable until you separate items requiring verbatim order from items requiring gist.
2. How would you redesign the post-call assessment to separate encoding failure from retrieval failure? Name at least three controls.
Read the worked answer
- Immediate versus delayed free recall: ask for the policy number within 10 seconds, then again after 5 minutes of filled delay. A large drop implicates storage or consolidation; stable but low recall implicates encoding. 2) Cued recall and recognition: offer digit groups ("was it 828 or 882?") and a four-alternative policy number. Recognition much better than free recall points to retrieval or strategy, not storage loss. 3) Dual-task control: repeat the call with and without concurrent typing, counterbalanced, to quantify the divided-attention cost. 4) Stimulus control: use recorded calls with matched accent and digit grouping, and measure background noise at the workstation. 5) Scoring rubric: score gist, verbatim order, and numerical accuracy separately. Without these controls the 4/10 cannot distinguish bad memory from bad listening conditions plus note-taking competition.
3. The trainer wants memorization drills. What is a more defensible training and accommodation plan, and what would you tell the trainer?
Read the worked answer
Drills target the wrong stage if the problem is encoding under load. A defensible plan: reduce concurrent load during critical capture by using a structured note template with fields for name, policy number in grouped digits, callback time, and reason, so the trainee offloads verbatim data instead of holding it in phonological storage. Teach a repeat-back protocol: "Let me read that back: 4471, 828, callback after 2:30." This converts a memory task into a verification task. Group digits into 3-4-4 chunks to match phonological capacity. Move the workstation or add acoustic treatment if 62 dBA is avoidable. Re-audit after two weeks with the same call set. Tell the trainer: the current evidence shows a specific encoding-under-divided-attention weakness and a reconstruction bias, not a stable trait; the 4/10 is confounded by typing and noise, so it should not drive a selection decision.