090Mathematics · Statistics

Statistics after a linear transformation

Statistics after a linear transformation

Statistics after a linear transformation · M=0: μ′,median′,range′,IQR′,v′,σ′; M=1 centre′; M=2 spread′; M=3 v′,σ′

  • 193 bytes
  • COMP mode
01

Worked examples

Example 1

A collection of measurements has population variance 9 square units. Every value is multiplied by 3 with no shift. Find the new population variance and standard deviation, then select the listed standard deviation.

  1. A.9
  2. B.81
  3. C.27
  4. D.87
Answer
A — v′=81 square units; σ′=9 units.
Standard solution (without a calculator)
  1. Identify the transformation: y=ax+b with a=3, b=0 and original v=9.
  2. Transform variance: v′=a²v=3²×9=81 square units.
  3. Take the square root: σ′=√v′=√81=9 units; equivalently |a|√v=3×3=9.
  4. Choose the option: The transformed standard deviation 9 is option A.
Do it on the calculator

Example 2

A set of distances has interquartile range 4 cm. Each value is transformed by y=−2x+7. Find the new interquartile range and select the matching option.

  1. A.4
  2. B.6
  3. C.8
  4. D.12
Answer
C — IQR′=8 cm.
Standard solution (without a calculator)
  1. Classify the statistic: IQR is a spread, so the shift +7 does not alter it.
  2. Use the magnitude of the scale: IQR′=|a|IQR=|−2|×4=2×4=8 cm.
  3. Check the sign: Reversing order with a negative coefficient still leaves IQR nonnegative.
  4. Choose the option: 8 cm is option C.
Do it on the calculator

Example 3

A dataset has mean 0.5, median 0, range 8, interquartile range 4 and population variance 8.25. Every observation is changed to y=0x+6. Find all six transformed summaries in display order and explain whether a standard score can be assigned to the transformed data.

Answer
μ′=6; median′=6; range′=0; IQR′=0; v′=0; σ′=0; standard score undefined.
Standard solution (without a calculator)
  1. Identify the constant map: a=0 and b=6, so every transformed observation equals 6.
  2. Transform centres: μ′=0(0.5)+6=6 and median′=0(0)+6=6.
  3. Transform spreads: range′=|0|×8=0, IQR′=|0|×4=0, v′=0²×8.25=0 and σ′=√0=0.
  4. Interpret: The displays are 6,6,0,0,0,0. Since σ′=0, z=(y−μ′)/σ′ is undefined for the transformed data.
Do it on the calculator
02

When to use it

Under y=ax+b, mean and median become a times the original value plus b. Range and IQR become |a| times their original values, population variance becomes a²v, and population standard deviation becomes |a|σ. A negative a reverses order but does not make spread negative.

Select the summary mode

  • Enter X as mode: 0 for all summaries, 1 for one centre, 2 for one spread, or 3 for population variance and standard deviation. Then enter A=a. Mode 0 asks B=b, C=mean, D=median, X=range, Y=IQR, M=population variance. Mode 1 asks B=b and C=one centre value; mode 2 asks C=one nonnegative spread; mode 3 asks M=nonnegative population variance.

Read the displayed results

  • Mode 0 displays transformed mean, median, range, IQR and population variance, then ends with population standard deviation. Mode 1 ends with the transformed centre; mode 2 with the transformed spread. Mode 3 displays variance, then ends with standard deviation. Press EXE after each paused value.

Question types

  • Use this for original HKDSE Mathematics Compulsory Part statistics questions about changing measurement units or rescaling data. State whether an input is a centre, spread or population variance; the program does not verify that separately entered summaries come from one dataset.

Before you start

  • The mode must be 0, 1, 2 or 3. An invalid mode deliberately gives Math ERROR; a may be positive, negative or zero.
  • In mode 0, enter all five original summaries from the same dataset. Range, IQR and population variance must be nonnegative. The program rejects a negative one with Math ERROR.
  • Mode 2 accepts a nonnegative spread measure, such as range or IQR, without b. Mode 3 accepts a nonnegative population variance, also without b. A negative input gives Math ERROR.
  • For a<0, mean and median use signed a, while range, IQR and σ use |a| and variance uses a². If a=0, all transformed values equal b and all transformed spreads are zero; their standard scores are undefined.
  • The program does not check whether supplied mean, median, range, IQR and variance describe a possible common dataset. Use consistent source units and distinguish population variance from sample variance.
  • At prompts, blank EXE may reuse a stored value. Enter intended zero explicitly. P1 uses 193 bytes of P1–P4’s 680-byte COMP space and overwrites X, A, B, C, D, Y, M; save any prior values.

Sign in to unlock the full programme

Checking access…

The programme code and appendix unlock together.

Checking access…

Sign in to unlock the full programme

Checking access…

The programme code and appendix unlock together.