Page 89 - Microsoft Word - 00 P1 IW Prelims.docx

Performance reports for management

Problems dealing with quantitative data

There are a number of common mistakes and misconceptions that people make
when using numerical data for performance measurement:

 Data sampling inappropriate – need a random sample representative of the
whole population.

 Failure to look at the underlying causes, e.g. a huge sales volume increase
driven by heavy discounting.

 Data processing inappropriate, e.g. a mean can be skewed by extreme
values.

 Poor presentation of data, e.g. a graph indicates dramatic changes due to the
scale choses.

 Inappropriate comparators or benchmarks, e.g. a 20% increase in year on
year sales is reported but this does not look so impressive when compared with
a 30% growth in the market.

 Failure to understand underlying samples, e.g. a 90% customer satisfaction
rating is misleading if only 10 out of 5,000 customers were questioned.

 Failure to understand percentages – quoting a percentage figure rather than
an absolute figure can be misleading, e.g. a 100% pass rate for a course does
not look so good in absolute terms if only two students sat the exam.

 Selective use of figures, e.g. a manager boasts about revenue growth but fails
to report a fall in profits.

 Confusing correlation and causation, e.g. a manager connects an increase
in revenue with an increase in investment but it is actually due to an economic
upturn.

79