January 2001
Volume 33 · Issue 1
Contents
Test on Quality Control Statistics
and Concepts
by Lloyd S. Nelson
INTRODUCTION
Nearly fifty years ago Swan and Hicks (1954) published a
list of thirty multiplechoice questions designed to test
the reader's knowledge of quality control statistics. On the
assumption that few presentday practitioners have seen this
article I have selected, and occasionally edited, twenty of
the questions for those who would like to test their "IQ
in QC." Answers are given in Table 1, but the question
numbers have been randomized to make it difficult for peeking
to pay off.
 If control charting shows that the number of nonconforming
items in various lots is in control, we can conclude that
 The manufacturing process was in control
 The product was well mixed before dividing into lots
 Either a or b is true
 All lots should be accepted
 Probability paper can be used
 To inspect a distribution for normality
 For control charting
 To reduce the importance of extreme values
 None of the above
 If drawing tolerances and the natural tolerances of a
process operating under the normal law are the same, then
the control limits for Xbar will be
 The tolerance limits
 The tolerance limits times n
 The tolerance limits divided by n
 The tolerance limits divided by _{}
 The sum of the deviations of a group of measures from
their mean divided by the number of measures equals
 _{}
 _{}^{2}
 Zero
 Xbar
 Xbar control charts to examine the variation between
averages of samples can be drawn by plotting
 Subgroup totals
 Subgroup ranges
 Subgroup sigmas
 Subgroup size
 MILSTD105 is used for
 Establishing control limits
 Setting tolerances
 Accepting material
 Inspecting threads
 A.O.Q.L means
 Average outgoing quality level
 Average outgoing quality limit
 Average outside quality limit
 Anticipated optimum quality level
 The probability of drawing at random exactly one nonconforming
unit in a sample of 5 from a lot of 50 containing 20 percent
nonconforming is _{}
divided into
 _{}
 _{}
 _{}
 _{}
 A p chart is used when data consist of
 Variables
 Attributes
 Standard deviations
 Weights
 In setting up a control chart for a certain process, sigma
is found to be 11 units. Repeated measurements of the same
part, using the same technique on which the chart is based,
show a standard error of measurement of 5 units. The best
estimate of the true standard deviation of the process is
nearest to
 6 units
 10 units
 11 units
 12 units
 The standard deviation of the Poisson distribution is
given by
 _{}
 _{}
 _{}
 None of these
 When A and B are distributed independently, _{A
– B} is equal to
 _{A}
+ _{B}
 _{A}
– _{B}
 _{}
 _{}
 When measurements show a lack of statistical control,
the standard error of the mean
 Is related to confidence limits
 Is a measure of process variability
 Is simple to compute
 Has no meaning
 If nothing is known concerning the pattern of variation
of a set of numbers, we can calculate the standard deviation
of the set of numbers and state that the sample mean ±3
times the calculated standard deviation will include at
least
 88.9 percent of the population
 95 percent of the population
 99.7 percent of the population
 100 percent of the population
 A block has a height specification of 1.000" ±0.015".
Three blocks are stacked to make anassembly. The difference
in height between the largest and smallest of 100 such assemblies
is nearest to
 0.030"
 0.060"
 0.090"
 More information is needed before an answer can be
given
 It is sometimes economical to permit a process that is
being monitored by an Xbar, R chart to go
out of control when
 Individual R's exceed Rbar
 Cost of inspection is high
 Six sigma is appreciably less than the difference
between specification limits
 The Xbar control limits are inside the drawing
tolerance limits
 A pchart based on samples taken from each box
of a large shipment of parts can be used to test
 The homogeneity of the shipment
 Whether or not the parts are produced under control
 Whether or not the parts are within specification
 None of the above
 As compared with a single point outside a 3_{}
limit two points in succession between a 2_{}
and a 3_{}
limit on an Xbar control chart for means are
 A more significant indication of an assignable cause
 An equally significant indication of an assignable
cause
 A less significant indication of an assignable cause
 Unimportant
 A cumulative frequency distribution is called
 An operating characteristic
 A histogram
 A frequency polygon
 An ogive
 A consumer's risk of 10% means that
 The probability that a sampling plan will reject "good"
material is 10%
 The probability that a sampling plan will accept "poor"
material is 10%
 The acceptable quality level of the lot is 10%
 The unacceptable quality level of the lot is 10%.
TABLE
1. Answers
Reference
Swan, R. O. and Hicks, C. R. (1954). "What's Your IQ
in QC". Industrial Quality Control Vol. X, No,
6, pp. 8487.
