r Appendix K Glossary 0l lerms and symbols

,4 Constants used in the calculation of the control line for mean, mov-
ing mean, median and mid'rang contol chart, with vadous suffixes.
Accrldcy Associa ted with the nearnes6 oI a process to the tatget value.
Action Limit
(line)
Line on a conhol chaft bevond which the Drobabil-
ily of tinding an observation is such that it indicates that a change ha'
occurred to the prccess and that action should be taken to investigate
and
/or correct for the change.
Aclion zoke The zones outside the action limits/lines on a control
chart where a result is a clear indication of the need for action.
,4RL The average run lentth to detection of a change in a process.
Assigtlable ca ses Sources of va-riatiori for which an explicit reason exist6.
Attrib te clhlts Control drarb used to assess the caDabilitv and monitor
the performance of parameters a666sed as attdbutes or discrete data,
Attribute doto DiscEte data which can be counted or classified in some
meaningful way which doeB not include measurement.
ALErage See Mearr,
B Corutants used in the calculation of conEol chart Iines for standard
deviation charts.
Bar A bar placed above any mathematical symbol indicates that it is
the mean value,
Bal chart Adiagtarnwhich represents the lelative frequency oI data.
Binomkl distribution A prcbability distribution for 6amples of attrib-
utes which applies when both the number of conforming and non-
conJorming items is known,
Braiflstormi g An activity, normally carded out in groups, in which the
participants are mcouraged to allow their experince and imagination
to run wild, while centred around specific aspects of a problem or effect.
c crrll A control chart used for attributes when the sample is constant
and only the number of non-conformances is l<nowni c is the symbol
which rpresents the number of non"conJormances present in samples
oIa conslant size. c-bar (-) reprcsents the average value of a series of
Copable A, process which is in statistical conhol and for which dre
combination of &e degree of random variation and the ability of the
442 Sl al sl i cal Pr ocess Conl r ol
a
control procedur to decc !l
of the specification.
Cause and
4Jcct
dias
-.
rt
tionship behveen an end:: d
CenLrcl tendency The J-.!-r
value,
&ntre line
(CL)
A lin .r r,r
Cft?c&lisl Alist used ro rEcl
Comfton causes Sea Rani.c!,a
Conhtfting Tota y ul.agtrq'
Conlifiuous ilata
Quantt-lE,
all measured values are
F<Edl
Corrhol The ability or nGi
-
data observed, interpret d.E.lt
if justified.
Conhol chtrt Agraphicat d
ily distinguish
between rar6
Cantrcl limits (lines)
Lim'E* f
the zones of stability (no
acu
and the ned to sek addibfld
Coufitable ilata A lorm oi fuc
can only b counted (s.,. r:i- {,t
Cp A process capabilih irrial
quency distribution to the b).1!
Cp* A process capabilih it*
quency digtribution and tl| rifi
width of the specification
Cusuft chart A graphic pr!:{d
chart is particularly sensiEr?
cnanges,
Cusurft scole The cumuianre *
of observed valus and a prl.d6
d, or d2 Symbols which .epre
between the standard deliatiE
D Symbol rr'hich rcpresnG d
trol limits on a range chart, rr:d
t
I APPend ces 443
control procedure to detect change is consistent with the requirements
of the specification.
Atuse and effect diagram A
Braphic
display which illustrats the rela-
tionship betwen an effect and its contributory causes.
Central lcwlency Th clustrin8 of a population about some prefered
Centn: line
(CL)
A line on a control chart at the value of the process
Creclllsl A list used to ensur that a ll steps in a prccdur are caftied ou t.
Commoncauses S?? Random causes,
Confotrting Totally )^ ageemmt with the specificafion or rquircments.
Continuaus data
Quantitative
data conccming a parameter in which
all measured values arc possible, even if limited to a specific range.
Corflol The ability or need to observe/monitor a process, record thc
data obsewcd, in terpret the data rccorded and take action on the process
ifjustified.
Conlral chart Agraphical method ol recording results in order toread-
ily distinguish between random and assignable causcs of variation.
Conlrcl linlits
(line
Limits or lines st on control charts whic:h scparate
the zones of stability (no action required). wahing (possible problems
and the need to seek additional informaiion) and action.
Countable dafi A form of discrete data where occu[ences or evcnts
can only be counted (sce dlso Attdbute data).
Cp A process capability index based on the ratio of the spread of a fre-
quency distribution to the width of the spccification.
Cpt A process capability indcx based on both thc ccntring of a frc-
quency distribution and the ratio of the spread of thc distribution b the
width of the spcification.
Cusunt chdrt A graphic pesentation of the cusum scorc. The cusum
chart is parti.ularly sensitive to the detction of small sustained
chanSes.
C sunt score The cumulativc sum of the differenccs between a series
ofobserved valus and a predetermined targt or average valu.
d, or d2 Symbols which represent Harttey's constant, rhe relarionship
between the standard deviation (d) and ihe mean range (R).
D Symbol which represents the constant used to determinc the con-
trol limits on a rangc chart, with various suffixes.
I
I l qq St at st i ca
process
Cont rol
Data FacE.
DeJecl A fa'Jlt or flaw which is not permitted by rhe specification
requirements.
D$ectiLv An ltem vthich contains one or morc defecb and/or is judged
to be non-conforming.
Defeclio, The act of discovering.
De?latio, The dispersion betwem two or mor data.
DifeNnce chart A control chart for differences fiom a target value.
Discrct? data Data not available on a continuous scale (9ee 4lso Attribute
data).
Dispelsio, The spread or scatter about a cmtral tendency,
DM,4IC Six sigma improvemmt model - Define, Measure, Analyse,
Improve, Conhol.
Frequency Ho\\t olte samething occurs,
Frcquency distributio.a A table or graph which displays how frequently
some values occur by comparison with others, Common distributions
include normal, binomial and Poisson.
Gnnil mean The mean of either a whole population or the mean of a
series of samples taken from the population. The grand mean is an esti-
mate ofthe true mean (sse Mu).
Histografi A diagram which represents the relative frequncy of data.
Individual An isolated, result or observation.
IndfuAuols plot A graph showing a set of individual results.
L,4L Lower action limit or line.
LCL Lower control limit or line.
LSL Lowff specification limit.
LWI Lower warning limit or line.
&R Median of the sample mid-ranges.
Meaa The average of a set oI individual results, calculated by adding
together all the individual rcsults and dividinS by the number o]
iesults. Means arc rpresented by a series of symbols and often carry a
bar above the symbol which indicates ihat it is a mea]l value.
aPr
Mod.
byr
step:
by*
steps
Mu( ,
of af
nn
ment!
requn
The n,
popt!r
value,
and 99
asamF
items a
r
I Appendi ces 445
Mun chatt Agtaphwith conrrollines used to monitor the accuracv of
a procss/ being assessed by a ptotofsample mans.
M.ar? r'dr.qe The mean ofa series of sample ranges.
Mu satnple sizt The average or mean of rh sample sizes.
Mcdid, The central value within n poputation
above and bIow which
there are an equal number of member; of the popularion.
Mode The most frequcntly occurring vatue within a population.
MaDi gt ean Am!'anvatue calculated from a sedes of individua I vatues
by moving the sampl for calcuiation of the man through the series in
sicps ofone individual value and without changing theiample size.
Mooittg ranY A range vatue calculr tcd from a seris of individ uai values
by moving ihe sample forcatcuhtion of the ranSe throuSh the scries in
steps of one individual value and wiihour chan;ing the'sample size.
Mll (/i) The Grcek letter used as the symbol to represcnt the true mcan
oi a population as opposed to the various estimntes of this value whjch
measLrrement and calculation make possjble.
, The number of i ndi vj dual s wi t hi n a sampl c of si ze rr. , _bar (n) i s
th averate size ofa series of sarhples,
Nan-conf)tnifig
Not in conformance with the specification/require_
Non-conl,rmititi Defects, etort faults with respect to the spccification
/
Nor wl dislribution Also known as the Gaussian distribution of a coniin_
uous variable and sometimes referred to as the
,bellshaped,
djstribution.
The normal disdiblrrion has the characteristic rhat 68.2; per ceni of thc
population
js
contained within 1 one standard deviation from thc mean
value, 95.45 per cent within a hvo standard deviations from the mean
and 99.73 per cent withhin 1 thrle standard deviations ftom rhe mean.
l]/ .rali A controi chart used for attdbutes when the sample siz is
constanl and the number of conforming and non-conJorming items wthin
a sample are bothknown. /r is th sampte size and pis the proportion of
non-contormtng rtems.
p':rdlt A control chaft used for attriburs showing the proporrion oI
non{onformins item. in d sample. p L the proporrion,rf
non_conJormin!
rt em, . rnd p-bJr rt ) r, . pre, enr. t he avcrrge of d seri e\ of vdt uc. ot p.
Parcta ana.lysis. A technique of ranking data in order to disthguish
btween the vital few and the trivial many.
I
I 446 St al i st i cat Process Conl roi
Poisson distib tion A probability diBtribution for samples oI attributes
which applies whn only ihe number of non-confomities is known.
Pop latiotl The tullset of data from which samples may be taken.
P/ecision Associated with the scatter about a central tmdency.
Prcpention The act of seekint to stop somethint occurrint.
Probability A rneaswe of the lilelihood of an occturence ol incident.
P/ocess Any actjvity which converts inputs into outputs.
Prccess capability A measure of the capability of a proc$s achievd by
assessing the statistical state of control of the process and the amount of
random variation ptesent. It rnay also refer to the tolerance allowed by
the specification,
Prccess c,pability indet An index of capability ( see Cp a A Cpkr.
Process control The management of a proce66 by observatio& analysis.
interprtation and action designed to limit variation.
Process mean The avetate value of an attribute or a variable within a
prccess.
Ptoportion thf?ctive The ratio of the defective! to the sample size,
rc,presented by the symbol p. p-bar (p ) rcpresenis the average oJ a sericu
or values otp,
Qlrlliy
Meeting the customer requirements.
R The range ofvalues in a sample.
R-rar (R) The symbol for ihe mean of a series of sample ranges.
F The median of sample ranges.
Randofi causes The contributions to vadation which are random in
their behaviour,
j.e.
not structured or assitnable.
Range (R) The difference between the largest and the smallest result in
a Eample of individuals - an approximate and easy measur oI the de$ee
oI scatter.
Rafige chort A gaph\Njth control lines used to monitor ihe precision
of a process, being assessed by a plot of sample ranges.
Rm A st of results which appea$ to lie in an orderd series.
R cirl, A graph wirh contlol lines used to plot individuat iesulrs.
Sahple
is often
of the w
Sample s
or the sil
Scattet
Scatter d
the other
Shewhart
posed by
Sigma
lo
a popula
Six sigtt
focussing
Stuwed di
about th
Special cal
SpeciJicoti
inputs or
Spnlad R
SQC Sta
on produ(
Stuble n
Stable zonl
chartand
Standad d
deviation
Standa el
ure of th
rePresente
Statistical cr
the origin (
a
T Append ces 447
SdrTph A group of individunl rcsults, obsenations or data. sample
is oftcn used for assessmnt with a licw to determining the properties
of thc whole population or univcrbc from which it is drawn.
Sn
rylcsize
(n)
The number of indiv ualresults included in a sarnple,
or thc sizc of the sample taken.
S.iitfcr Rcfers to the dispersion of a distribution.
Scattrt Lliagrnm Th picturc {,hich results *'hen simultancous resulis
for h\rc varyint paramtcrs are plotted together/ one on thc r axis and
the other on the
v
axis.
Sh.lrlhart
(hltts
The controlchnrts for a ttrib utes and varhb lcs first pro
posed by Shewhart. Thcsc inclLrde mean and range, ,p, p,. and r/ charts.
sn,'d (d) The Creek leitr used k) signify the standard dcviati()n of
-Sijr
sLg , A disciplined approach for improving performance bv
focussinS on producingbetter products nnd soviccs faster and chcnp.'r.
Skttrcd dislribuliott A frcquency distribution r41ich is not symmerrical
about t hc mc. rn v. l ue.
SPC Scc Stntistical process coitrol.
Sfr0nl fnrs(N Scc Assignablc causcs.
Sl)cct1.'rtuirr Thc relluirement against which thf acceptability of thc
inputs or ouiputs of a process aro k) b(' judgod.
Spard lt{rfcrs to the dispersion ofn distribution.
SQC Statisticalquality control - similar to SPC but with an emphasis
on product qu.ility and less mphasis on proccss onirol.
St,b/. The term used iodescribca proccss when noeviLlence oa.ssign
ablc causes is present.
Stdbl.'rott' The central zone between tho warning limiis on a control
chart and within which most of ihc rcsults are expected to fall-
Stnru1afl1 df i|tion (o) A measufe of thc spread or scatter of a popula-
tion around its central tendency. Various cstimates of the standard
deviation air ropresented by symbols such as d, diri
r)
and s.
Standd lel )r The standard deviation ofsample meanvtrlues-a mcas-
urc oi their spread or scatter around the grand or process mcan,
rcpresented by the symbol SE (or di).
Statisti?l canttol A cond ition des{dbing a prGess for which th observed
values are scattefed about a mean value in such a way as to imply that
theoriBinofthe variaiions is entirely random with no assignable causes
ol variation an.l no runs or trends.
I
I cl a St al r st i cal Pr ocess Conl r ot
Statistical placcss contlal The us of sratistically based rechniques for
the control of a process for hansfomint inputs into outputs.
Sfallsti.s The colleciion and use of data - methods of distillin8 infor-
mation from data.
f The valu of d statistic calculated to test the significance of the djf_
frcnce betwen haro means.
T Asymbolused to represnt a toterance limit (1?).
I rl / ui 4r4 A. i mpl ct o, , l f or re(
^rdLnF,
e\ (nt s ds t he) o. curort oe\ f ract
t rcquenci es f r, )m \ i st i nt s l i srs or dat a
ftrycf The objective ro be achived and againsr which pcrformance
will be assessed, often thc mid-point of a specification.
Tblerance The d,lffercnce between the lowest and/or the highest value
stated in the specification and the mid-point of the specification.
?erd A series of rcsults which show an upward or downward tendency.
ll crarf Acontrolchart used for attributes when the sample size is not
constant and only the number of non-conformities is kn;wn. ,1 is the
symbol wfuch represents the numbr of non-conformities found in a
single sample and !-bar (t) rpresents the meanvaluc of r.
UAI Upper action limit or line.
IJCL Upper control lirhit or line.
llri.,e/se SrePopulation.
USL Upper specification limit.
UIVL Upper waming limit or line.
V-ftarft
_A
device used in conjunction with a cusum chart to identify
trends of known significance.
Varitble data Data which is assessed by measurement.
Vatianca A measure of spread equal to the standard deviation
squarcd (d:).
Variatiotl The inevitable differences btweer ourpurs.
Waming Linit
(liae)
Lines on a control cha4 on each side of the cenhal
Iine- and within which most results are expected to fatl but beyond
which the probabiljty of finding anobseffation is such that it should be
regarded as a waming of a possible problem
X
( r
, \ :
X-r
i
X
Z
z.
l
i&
l \ ' r r r , r , . r r r , I l r ( / r nx, s ( r r
. r . r ! r l f , ) l
n. Li or r l , , i i l s, ) nl \ i t l r i n r \ l r l , h I r . \ L, l l
ApNei d r r s i t Jl t
L r r r r t l r ct $1, f n t l r . r ' , r r ) r r r ! . r r r l I l j ( ,
{r i t gt \ i \ t r ,
l r Nsr l r i l
l \ r n , r L. h, I r , {,
r , \ ' r i f t l i \ i ( l L/ , r ! . r l uf , ) l , r \ , , i r . r l j l ( .
I i r r r \ r l l x l r ( . , 1r \ . r ! ( , ( Jl , . . , , I j l r t ( . , j ) nr ( . t , , , , , \
t r . . . \ j j r t r j \ l , , r l
\ / r , / , 1' , r r \ r l h| r ; r . | r r i or
l ,
r \ ( . r r r i | . : ( f I 1. r 1, , , , , . I r . r \ r ) j t r j
I l l r . nr ( \ l , , i \ , r Li , . , , r , i : , | | 1, 1.
\ l hr gr , r r r , l or
l , r \ t , s. . , , r . LI i , r
I \ i l r ( . , i l , r \ i r j r f t , .
, / l l , ( . : l , l r l . r , l r 1t \ l , i , , r i , r \ , i , . r t ( . t r . | | nr l ! . , , , t : t , r , r Li , r n, t , , \ i , l
l r i r r s l ! , 1s( r . I l h( noj r r . t , j l j j r l , j l Ll
. .
r I r I ) I I L
j ( I I . I r I , I , l l j t t j ( . r , t ( r ! \ l
\ , r 1, 1. , , ! ( I r , I l r f l r .
/ i r r r l . \ r t r r l f , , l ( l r , r . t ( f . , L, I r (
l , l i l ( \ i ( t j i t , . r . l ( t - . l , i , nr , r t , , j . L\ , , t ( ( ,