Devanagari typewriting error cluster on ra-conjuncts
In a 10-minute Marathi typewriting test on a Remington layout, a candidate types 300 words with 31 errors, giving a net speed of 28 wpm. The error sheet shows 18 errors in ra-conjuncts (क्र, प्र, ट्र, द्र), 7 in matra placement (ि versus ी, ु versus ू), 4 punctuation errors, and 2 skipped words. The candidate looks at the keyboard when typing conjuncts and practises by typing long passages without isolated drills. The target is 40 wpm with 95% accuracy in three weeks. The candidate asks whether to type 15 minutes more each day to build speed or to stop and fix accuracy first. The test uses Unicode Devanagari, and the candidate often uses the mouse to correct errors.
1. Should the candidate prioritise speed or accuracy in the next three weeks? Justify with the error rate and the exam's likely scoring.
Read the worked answer
Prioritise accuracy first. The candidate is not near a speed plateau; the speed is being generated by inaccurate keystrokes.
Calculation: 31 errors in 300 words is about 10.3 errors per 100 words. At 40 wpm the candidate would type 400 words in 10 minutes; at the same error rate that is about 41 errors, which will almost certainly fail a 95% accuracy requirement. Net speed is already reduced to 28 wpm, so the apparent 30+ gross speed is misleading.
The error cluster is also mechanical, not lexical: 18 ra-conjunct errors and 7 matra errors show that the candidate has not automated the Devanagari key combinations. Typing more long passages will rehearse the same errors. The correct sequence is: isolate the high-frequency conjuncts, build accuracy at a slower speed, then raise speed. Use the mouse only after a timed section, because mouse correction breaks touch-typing rhythm and hides the true error rate.
2. Design a 20-minute daily drill that targets ra-conjuncts and matra placement while preserving touch-typing discipline.
Read the worked answer
Daily 20-minute block:
- 5 minutes: home-row and matra warm-up. Type ि/ी and ु/ू in minimal pairs (कि/की, कु/कू) without looking at the keyboard. Accuracy target: 98%.
- 7 minutes: ra-conjunct drill. Type क्र, प्र, ट्र, द्र in syllables and then in words (प्रकरण, क्रम, ट्रक, द्रुत). Start at 15 wpm and increase only when two consecutive rounds are error-free.
- 5 minutes: two-minute sprint at a comfortable speed. Count errors per 100 words; do not correct with the mouse during the sprint.
- 3 minutes: review the error sheet and retype only the mistyped words.
Touch-typing discipline: cover the keyboard or use a keyboard mask for the first 10 minutes. If the candidate must look for a conjunct, stop, locate the key by feel, and repeat the word five times. Accuracy is measured by errors per 100 words, not by how fast the drill felt.
3. What evidence should the candidate use before raising the target from 28 to 40 wpm?
Read the worked answer
Use three evidence points, not a single good day:
- Accuracy stability: three consecutive 5-minute tests at 30–32 wpm with at least 97% accuracy and no more than 3 errors per 100 words.
- Error distribution: ra-conjunct and matra errors should fall below 1 per 100 words each. If they remain the dominant errors, the speed increase is premature.
- Touch-typing behaviour: the candidate should type a full two-minute passage without looking at the keyboard and without mouse correction. If either occurs, the speed is being borrowed from visual checking, not from automatic motor memory.
Once these hold, raise the target in 2 wpm steps, not directly to 40. A direct jump will likely reproduce the same 10% error rate at a higher gross speed, which lowers net speed and wastes test time.