A normal distribution is unbounded and may assign non-zero probability outside the possible score range; tail behaviour may also not match real assessment data.
Part (a): The 92nd percentile of Z is z≈1.405.
Part (b): Cutoff=68+1.405(12)=84.86, so about 85 marks if scores are whole numbers.
Part (c): A normal distribution is unbounded and may assign non-zero probability outside the possible score range; tail behaviour may also not match real assessment data.
Editorial practice rubric; not an official QCAA marking scheme. Equivalent correct methods are accepted. Method marks may be awarded after an earlier arithmetic error when the subsequent method is valid.
Part (a): Identifies the lower-tail probability 0.92 for the upper 8% cutoff.
[1 mark]
Part (a): Uses inverse normal technology to obtain z≈1.405.
[1 mark]
Part (b): Converts using x=68+12z.
[1 mark]
Part (b): Obtains approximately 84.86; a cutoff of 85 is accepted when integer scores are specified.
[1 mark]
Part (c): Explains a relevant limitation, such as the unbounded normal model assigning probability outside the possible score range, or its tails not matching real scores.