Block 4 - lecture 4

Question 1

aspects of charts for variable subgroup size?

Answer 1

• different control limits
• higher n value narrows the limits
• difficult to interpret
  SHOULD BE AVOIDED

Question 2

centreline of charts for trends?

Answer 2

equation of centreline can be calculated using:
Xbar = a + bG

(equations for a and b in bklet, make a table with G, Xbar, GXbar and G^2)

Question 3

control limits for charts for trends?

Answer 3

A2 Rbar or A sigma subscript 0 either side of the centerline.

Question 4

charts for moving average?

Answer 4

average is taken of the previous n number of samples

Question 5

charts for moving range?

Answer 5

range is taken of the previous n number of samples

Question 6

Charts for median and range?

Answer 6

data is collected in subgroups like Xbar and R, but instead of a mean a median (Md) is taken.

Question 7

How to find the centerline of median and range charts?

Answer 7

find the median of the medians of the subgroups (Md subscript Md) and the median of the ranges (R subscript Md)

Question 8

How to find the control limits of median and range charts?

Answer 8

use specific equations in bklet

Question 9

advantages of charts for median and range?

Answer 9

• less arithmetic
• easy understanding
• easily maintained by operators`

Question 10

disadvantages of charts for median and range?

Answer 10

doesn’t take extreme values into account

Question 11

reasons only one measurement might be made in a sample?

Answer 11

• too expensive
• too time consuming
• destructive
• very few samples

Question 12

working out the range for charts for individual values?

Answer 12

find the range from the last point ie.2nd-1st

this is plotted between the subgroup numbers

Question 13

advantages of X chart?

Answer 13

• no calculations
• easy to do on shop floor
• direct comparison to specs

Question 14

disadvantages of X chart?

Answer 14

• requires a lot of subgroups to show out of control condition
• doesn’t summarize the data as well as Xbar

Question 15

Non-conformance limits?

Answer 15

a limit for Xbar, like specs are to individual values

Question 16

exponential weighted (geometric) moving average?

Answer 16

moving average that gives more weight to more recent values

Question 17

advantage of exponential weighted (geometric) moving average?

Answer 17

• can detect small shifts in the process faster than regular running averages

Question 18

disadvantage of exponential weighted (geometric) moving average?

Answer 18

• doesn’t react as quickly as the X bar chart

Question 19

weighting of current subgroup (lambda)?

Answer 19

between 0.05 and 0.25

smaller values detect change faster