Capstone: a full practice exam and career planning

Three weeks before her exam date, Kiran sat the study group's full practice exam on a Saturday morning in the customer lounge, with Marty as invigilator: 125 questions written by the group and by Marty over the winter, weighted exactly like the real exam, four hours on the clock, phones in the office. Theo sat a shorter version for his second-period exam. Bev brought coffee at the start and left them to it.

This capstone follows what Kiran did with the result: scoring by area, sorting her errors by reason, aiming her last weeks at what the result showed, the exam itself, and the plan she made for the years after certification. It brings together every lesson in the course and every course in the program.

Scoring by area

Kiran answered 104 of 125, which is 83 per cent and a clear pass. The score by major work activity was more useful than the total. She did well on her old strengths and on the areas she had worked hardest on, and one area stood out at the bottom.

# Score a practice exam by major work activity and find the weakest areas
results = {"A": (9, 8), "B": (22, 19), "C": (12, 8), "D": (17, 13), "E": (23, 19), "F": (23, 21), "G": (10, 9), "H": (9, 7)}
total_q = sum(q for q, c in results.values())
total_c = sum(c for q, c in results.values())
print(f"overall {total_c}/{total_q} = {total_c / total_q * 100:.0f}%  (pass at 88)")
ranked = sorted(results.items(), key=lambda kv: kv[1][1] / kv[1][0])
for area, (q, c) in ranked[:3]:
    print(f"study next: {area}  {c}/{q} = {c / q * 100:.0f}%,  {q - c} marks available")

overall 104/125 = 83%  (pass at 88)
study next: C  8/12 = 67%,  4 marks available
study next: D  13/17 = 76%,  4 marks available
study next: H  7/9 = 78%,  2 marks available

Sorting the errors

Kiran's error log for the practice exam had 21 entries. She sorted them by reason, as lesson 2 taught. Eight were knowledge gaps, five of them in network diagnosis. Seven were reasoning errors, almost all of the same kind: she had chosen a correct but unusual cause over the common one. Four were misreads, two of them missed the word except. Two were careless slips in calculations, one a unit error between kilopascals and megapascals and one a torque wrench setting where she multiplied by the adapter ratio instead of dividing.

# Error log summary and the expected gain from fixing each reason
log = [("knowledge", "network")] * 5 + [("knowledge", "other")] * 3 + [("reasoning", "mixed")] * 7
log += [("misread", "mixed")] * 4 + [("slip", "calculation")] * 2
from collections import Counter
by_reason = Counter(r for r, a in log)
recover = {"knowledge": 0.6, "reasoning": 0.5, "misread": 0.9, "slip": 0.9}   # share likely recovered
gain = 0
for reason, n in by_reason.most_common():
    g = n * recover[reason]
    gain += g
    print(f"{reason:<10}{n:3}  likely recovered {g:4.1f}")
print(f"expected extra correct answers: about {gain:.0f}")

knowledge   8  likely recovered  4.8
reasoning   7  likely recovered  3.5
misread     4  likely recovered  3.6
slip        2  likely recovered  1.8
expected extra correct answers: about 14

The arithmetic made the plan obvious. The misreads and slips were almost free: six marks for slowing down and writing units. The network gap was the biggest single block of lost marks and the one most likely to hurt on the real exam, where a few unlucky questions could push the area lower. And the reasoning errors needed a habit, not more facts.

The last three weeks

Kiran spent the last three weeks on three things. Two evenings went to network diagnosis on the shop's training board, measuring terminating resistance, reading CAN signals on the scope and finding a deliberately open CAN low wire that Marty had set up. Every study session ended with twenty mixed practice questions, read twice, with the most-common-cause question asked before choosing. And she wrote one last timed set of 50 questions under exam conditions a week before the exam, then stopped studying new material, reviewed her error log and her spaced-repetition cards, and slept.

# Pacing on exam day: checkpoints against the clock
total, minutes, review = 125, 240, 40
pace = (minutes - review) / total
for hour in (1, 2, 3):
    print(f"after {hour} h: about question {hour * 60 / pace:.0f}")
print(f"pace {pace:.1f} min per question, leaving {review} min for marked questions")

after 1 h: about question 38
after 2 h: about question 75
after 3 h: about question 112
pace 1.6 min per question, leaving 40 min for marked questions

Exam day

The exam was at the provincial apprenticeship office on a Tuesday morning in June. Kiran used the three-pass plan from lesson 2, reached question 40 at the end of the first hour, marked 18 questions for her second pass and had 45 minutes left for review. Two of her marked questions were about CAN bus resistance readings and she answered both with confidence. Her result arrived in her online account the following week: 86 per cent, a pass, with every major work activity above 70 per cent. Her journeyperson certificate with the Red Seal endorsement followed. Theo passed his second-period exam the same month.

Planning a career

Certification is a start, not an end. The night after her result, Marty took Kiran for dinner and asked her to write down three things: what kind of work she wanted to be doing in five years, what she would need to learn for it, and what it would earn. They talked through the paths the trade offers.

# Comparing paths over ten years (illustrative pay index, journeyperson today = 100)
paths = {"stay general": [100, 102, 104, 106, 108, 110, 112, 114, 116, 118],
         "diagnostic specialist": [100, 100, 108, 112, 116, 120, 124, 128, 132, 136],
         "EV specialist": [100, 98, 106, 112, 118, 124, 128, 132, 136, 140]}
for name, idx in paths.items():
    print(f"{name:<22} year 5: {idx[4]:3}  year 10: {idx[9]:3}  ten-year total: {sum(idx):5}")
print("training years can dip while study takes time; the gap opens later")

stay general           year 5: 108  year 10: 118  ten-year total:  1090
diagnostic specialist  year 5: 116  year 10: 136  ten-year total:  1176
EV specialist          year 5: 118  year 10: 140  ten-year total:  1194
training years can dip while study takes time; the gap opens later

Pay in the trade varies by province, by employer and by how a shop pays: hourly, flat rate or a mix. Kiran's notes used an index rather than figures, because her own wage would depend on Ridgeline's pay plan and on her own productivity. What the index showed was the shape of the choice: specialization usually costs a year or two of training time and pays back over the following decade, and the electrified and driver assistance work was growing fastest at Ridgeline.

Keeping certification current

A journeyperson certificate does not expire, but the trade keeps changing, and the knowledge behind the certificate does. Kiran's plan was to complete the maker high-voltage training Ridgeline's suppliers offered within a year, add a driver assistance calibration course, and keep her own record of training hours the way she had kept her apprenticeship record book, because employers and future courses would ask for it.

# Apprenticeship and training hours record
periods = {1: (1530, 9), 2: (1545, 9), 3: (1528, 9), 4: (1534, 9)}   # (hours on the job, weeks of school)
total_h = sum(h for h, w in periods.values())
total_w = sum(w for h, w in periods.values())
print(f"on-the-job hours: {total_h:,}  technical training: {total_w} weeks")
after = [("maker high-voltage course", 24), ("ADAS calibration course", 16), ("network diagnostics workshop", 8)]
print("planned training after certification:")
for name, h in after:
    print(f"  {name:<30}{h:3} h")

on-the-job hours: 6,137  technical training: 36 weeks
planned training after certification:
  maker high-voltage course      24 h
  ADAS calibration course        16 h
  network diagnostics workshop    8 h

What the program built

Looking back at the record book, Kiran could trace the whole program in her own work. AUTO-110's first oil service and lift checks. AUTO-120's timing chain. AUTO-130's seized caliper and corroded rotors. AUTO-210's pothole season. AUTO-220's parasitic draw. AUTO-230's manifold leak and AUTO-330's no-code hesitation. AUTO-310's pinion angle. AUTO-320's cottonwood-packed condenser. AUTO-410's isolation fault and AUTO-420's glass bracket. Each was a lesson in the same method Marty had written on her first day: don't guess; look it up, test it, then fix it. The exam asked about all of it, and the method was what answered.

Marty's last advice was about the next apprentice. Ridgeline was hiring a first-year in the fall, and Bev had asked Kiran to be the new apprentice's mentor. "The best way to keep what you know is to teach it," he said. "And they'll ask you things you never thought to ask."

What the group learned

Looking back over the season, the study group wrote down the habits that had made the most difference, so that the next group at Ridgeline could start with them. Practise under exam conditions, not in comfort. Score every practice set by area. Log every error with its reason. Spend the most time where weight and weakness meet. Read every stem twice. Answer every question. Ask what is most common before choosing what is possible. And sleep.

Theo added one of his own: study with someone who knows a different part of the trade. His tires and Kiran's electrical had covered each other's gaps all winter, and Marty's experience filled in what neither of them had seen. The group's question bank, about four hundred original questions by the end, went into the shop's training binder for the next apprentice.

Bev's view of certification was practical. A certified technician is easier to sell to customers, easier to insure and easier to trust with the shop's most difficult work, and a technician who keeps learning is the one who stays when the trade changes. She had already set aside next year's training budget for Kiran's electrified vehicle course and for Theo's next two tests.

The last entry in Kiran's record book from her apprenticeship was the one Marty asked her to write on the day she got her result: the date, the result and the name of the next apprentice she would train. The record book went into the shop's binder beside the question bank, where the next group would find it.