FINDING Z- SCORE, PERCENTILES AND

QUARTILES, AND COMPARING

STANDARD DEVIATION
M. MAIKEY ZAKI BIA
 MEASURES OF RELATIVE STANDING

• MEASURES OF RELATIVE STANDING :COMPARING MEASURES BETWEEN OR WITHIN DATA SET.

• Z SCORE: THE NUMBER OF STANDARD DEVIATION A SPECIFIC DATA VALUE (X) IS AWAY FROM MEAN.
X X
SAMPLE POPULATION
x-x
𝑋−𝑋 𝑋−𝜇 10 24 34 24 58
𝑍= 𝑍=
𝑆 𝜎 8
S

• ALLOWS COMPARISON OF THE VARIATION 3 Std.Dev. 3 Std.Dev.

IN TWO DIFFERENT SAMPLES OR POPULATIONS # of standard deviation

99.7% of data lies from the mean
 EXAMPLE
WE GOING TO COMPARE THE HEIGHT OF TWO DIFFERENT PEOPLE, FOR EXAMPLE LET IT BE
1. LYNDON B JOHNSON HEIGHT → 76’ (6.4 FOOT *12=76’)
• MEAN HEIGHT FOR PRESIDENTS → 71.5’
• STANDARD DEVIATION → 2.1’

2. SHAQUILLE O'NEAL’S HEIGHT → 86’ (7.2 FOOT*12=86’)

• MEAN HEIGHT FOR MIAMI HEAT(BASKET BALLPLAYERS)→ 80.0’
• STANDARD DEVIATION →3.3’

• THE QUESTION IS WHO IS RELATIVELY TALLER?

• WE WILL SAY SHAQUILLE O'NEAL IS OF COURSE TALLER!
• BUT RELATIVELY COMPARING TO ITS OWN POPULATIONS WHO IS TALLER? WE ARE CONSIDERING ALL
PRESIDENTS AND ALL MIAMI HEAT PLAYERS.

• (USE Z-SCORE FOR POPULATIONS)

• NOTE// WE ALWAYS TAKE THE ACTUAL (X) –(X) IF POSITIVE IS GREATER AND IF NEGATIVE WE GET LESS. WE
NEVER DO THE OPPOSITE WAY
 SOLUTION
1. LYNDON B JOHNSON HEIGHT → 76’ (6.4 FOOT *12=76’)
• MEAN HEIGHT FOR PRESIDENTS → 71.5’
• STANDARD DEVIATION → 2.1’
𝑋−𝜇
𝑍= 𝜎
76−71.5 4.5
𝑍= = ≈ 2.14 →UNUSUAL (WE HAVE CALCULATED HOW MANY STANDARD DEVIATIONS WE ARE AWAY FROM
2.1 2.1
MEAN) MEANS WE ARE 2.14 STANDARD DEVIATIONS AWAY FROM MEAN.
• THE ABOVE IS USUAL OR UNUSUAL? IT IS UNUSUAL BECAUSE IT IS ABOVE 2.

2. SHAQUILLE O'NEAL’S HEIGHT → 86’ (7.2 FOOT*12=86’)

• MEAN HEIGHT FOR MIAMI HEAT(BASKET BALLPLAYERS)→ 80.0’
• STANDARD DEVIATION →3.3’
𝑋−𝜇
𝑍=
𝜎
86−80.0 6
𝑍= = 3.3 ≈ 1.82 →USUAL
3.3
NOW WE SEE NOW WHO IS TALLER IS SHAQUILLE, BUT WHO IS RELATIVELY TALLER?
ACCORDING TO MIAMI HEAT ALL PLAYERS THEY ARE NOT THAT TALLER, BUT HE IS TALLER THAN MOST OF OTHER PLAYERS, BECAUSE
THE RESULT IS POSITIVE Z-SCORE.
BUT (SHAQUILLE) WAS NOT AS MUCH TALLER THAN HIS PLAYERS, THAN THE (LYNDON B JOHNSON) WAS FROM OTHER PRESIDENTS.
(LYNDON B JOHNSON) WAS MUCH TALLER THAN OTHER PRESIDENTS (2.14)
HOW TO TELL A USUAL FROM UNUSUAL WHEN TALKING ABOUT Z-
SCORE
WHO REMEMBER EMPIRICAL RULE?
EMPIRICAL RULE SAYS HOW MANY PERCENTAGE OF DATA FEEL IN :
• 1 STANDARD DEVIATION ? 68%
• 2 STANDARD DEVIATION ? 95%
• 3 STANDARD DEVIATION ? 99.7%

WHAT Z-SCORE GIVES YOU?

HOW MANY STANDARD DEVIATIONS WE ARE AWAY FROM THE MEAN.
• IF Z-SCORE =1 MEANS STANDARD DEVIATION IS 1, IF Z-SCORE =-1 MEANS STANDARD DEVIATION IS
1IN THE OTHER DIRECTION. IF IT IS BETWEEN 1 & -1 MEANS YOU ARE ONE STANDARD DEVIATION
AWAY, MEANS 65% BETWEEN 1 &-1. IT IS THE SAME INFORMATION BUT SLIGHTLY IS GIVEN IN
OTHER WAY.
• WHAT IS Z-SCORE IF I GIVE YOU THE VALUE OF MEAN ITSELF? WHAT IN PREVIOUS EXAMPLE L.B.J
WAS 71.5 AS WELL WE WERE GOING TO GET ZERO, SO Z-SCORE AT THE MEAN IS ZERO.
 Z-SCORE IS ZERO
• A Z-SCORE BETWEEN -2 AND 2 IS CONSIDERED USUAL
• UNUSUAL LESS THAN -2 OR GREATER THAN 2.

• Z-SCORE TELLS US HOW USUAL AND UNUSUAL OUR DATA IS.

• WILL WE SAY (SHAQUILLE) HAS A USUAL HEIGHT OR
UNUSUAL HEIGHT?
• (SHAQUILLE) HAS A USUAL HEIGHT
• (LYNDON B JOHNSON) WE SAY UNUSUAL,
• WE SAY IT IS CLOSE TO BE UNUSUAL, NOT VERY UNUSUAL. IF
WE SAY 3 WILL BE VERY UNUSUAL. AND 4 WILL BE A GIANT!
OH! THAT WILL BE VERY UNUSUAL HEIGHT FOR A PRESIDENT.
FOR HEIGHT WISE HE WILL BE VERY TALL FOR A PRESIDENT.

• IT ALSO TELLS US HOW RARE A PIECE OF DATA IS 95%

-3 -2 -1 1 2 3
(THIS INFORMATION WILL BE VERY USEFUL FOR HYPOTHESIS TO
SEE IF IT IS TRUE OR FALSE)
 EXAMPLE
• WHAT IF (L.B.J) PLAYS FOR THE MIAMI HEAT? WIL THAT BE USUAL OR UNUSUAL?
• LYNDON B JOHNSON HEIGHT → 76’ (6.4 FOOT *12=76’)
• MEAN HEIGHT FOR MIAMI HEAT(BASKET BALLPLAYERS)→ 80.0’
• STANDARD DEVIATION →3.3’
𝑋−𝜇
𝑍=
𝜎
76−80.0 −4
𝑍= = ≈ −1.21 →USUAL
3.3 3.3

• IT WILL BE USUAL IF (L.B.J) SUCH HIS HEIGHT TO PLAYS FOR THE MIAMI HEAT. BECAUSE IT LIES
WITH THE RANGE.
 EXAMPLE
• ME.L ON MIAMI HEAT, LETS CONSIDER THAT THE HEIGHT OF (ME.L)→72

𝑋−𝜇
𝑍=
𝜎
72−80.0 −8
𝑍= = ≈ −2.42 →UNUSUAL..(IT MEANS IT WILL BE COMMONLY RARE, BUT NOT
3.3 3.3
IMPOSSIBLE TO BE IN TEAM WITH THAT HEIGHT)

THE LARGER Z-SCORE IS, THE MORE RARE THE PIECE OF DATA IS.
 RARENESS OF THE DATA

• THE LARGER THE Z-SCORE, IN TERMS OF ABSOLUTE VALUE THE RARER THE DATA VALUE.
• WHAT IS THE RELATIONSHIP BETWEEN RARENESS OF THE DATA AND THE Z-SCORE?
• DO YOU KNOW HOW TO CALCULATE Z-SCORE?
 QUARTILES
• IT BREAKS YOUR DATA INTO QUARTERS, IT BREAKS YOUR DATA EVERY QUARTER.
• IT IS VERY SIMILAR TO THE MEDIAN, THE MEDIAN BREAKS YOUR DATA INTO MIDDLE!

1ST QUARTILE: Q1 IT REPRESENTS THE BOTTOM 25% OF THE DATA, BUT IS THE DATA WILL BE ORDERED OR NOT
ORDERED? WHAT YOU THINK! IT SHOULD BE ORDERED RIGHT. SO Q1 REPRESENTS THE LOWEST 25% OF THE DATA
VALUES.
• Q1 → BOTTOM 25% OF SORTED DATA

2ND QUARTILE:
Q2 (MEDIAN 𝑴) → BOTTOM 50% OF SORTED DATA

3RD QUARTILE:
Q3 → BOTTOM 75% OF SORTED DATA

WE ARE CUTTING OUR DATA INTO 4 PIECES SO WE NEED HERE QUARTILES, THERE ARE NO 4TH QUARTILE
WE UST NEED Q1 , Q2 & Q3
 EXAMPLE(QUARTILES-1)
• 1, 3 , 6 , 10, 15, 21, 28, 36
ANSWER:
• SO WHAT IS THE FIRST THING WE LOOK AT THE DATA FOR? WE NEED TO SEE IF IT IS SORTED?
• GO AHEAD AND SEE IF YOU CAN FIND THE MEDIAN?
• 1, 3 , 6 , 10, 15, 21, 28, 36 M=?
 EXAMPLE(QUARTILES-1)..CONT
• 1, 3 , 6 , 10, 15, 21, 28, 36
ANSWER:
• SO WHAT IS THE FIRST THING WE LOOK AT THE DATA FOR? WE NEED TO SEE IF IT IS SORTED?
• GO AHEAD AND SEE IF YOU CAN FIND THE MEDIAN?
10+15 25
• 1, 3 , 6 , 10, 15, 21, 28, 36 →M= ⇒ ⇒
2 2
• M=12.5

• NOW WE FOUND THE 2ND QUARTILE WHEN WE FIND THE MEDIAN. FIND THE MEDIAN BEFORE OF
LEFT YOU CALCULATE 1ST QUARTILE, CALCULATE MEDIAN ON THE RIGHT YOU CALCULATE THE 3RD
QUARTILE

• 1, 3 , 6 , 10, 15, 21, 28, 36

1ST Q1 2ND Q2 3RD Q3

 EXAMPLE(QUARTILES-1)..CONT
• 1, 3 , 6 , 10 , 15 , 21 , 28 , 36

1ST Q1=4.5 2ND Q2=12.5 3RD Q3=24.5

 EXAMPLE(QUARTILES-2)

1 , 3 , 6 , 10 , 15 , 21 , 28 , 36 , 39
 EXAMPLE(QUARTILES-2)

1 , 3 , 6 , 10 , 15 , 21 , 28 , 36 , 39

ANSWER:
• 1 , 3 , 6 , 10 , 15 , 21 , 28 , 36 , 39
• WHEN WE CALCULATE THE QUARTILE HERE WE DON’T ADD THE 15 WHEN WE CALCULATE THE
OTHER QUARTILES. MEAN WE DON’T ADD 15 WITH CALCULATING THE FIRST AND THIRD
QUARTILE, WE JUST EXCLUDE THE 15 FROM THE CALCULATIONS.
• WE IMAGINE THERE IS NO 15 WE JUST SPLIT OUR DATA INTO 50%
1 , 3 , 6 , 10 , 15 , 21 , 28 , 36 , 39
•
• Q1=4.5 M=15 Q3=32
 PERCENTILES
• PERCENT IS OUT OF 100.
• PERCENTILES: IT SEPARATES THE DATA INTO 100 PARTS (JUST LIKE QUARTILE WHICH
SEPARATES 3 Q, BECAUSE WE WANTED 4 PARTS) . THEREFORE THERE ARE 99 PERCENTILES.

• THERE PERCENTILES TELL YOU WHERE YOU ARE, FOR EXAMPLE IF YOU TOOK 50% IN A TEST. IT
MEANS THAT YOU SCORE 50% BETTER THAN OTHER STUDENTS. AND NOT 50% ON THE WHOLE
TEST. IT MEANS IT COMPARES YOU WITH EVERYONE ELSE.

𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑣𝑎𝑙𝑢𝑒𝑠 𝑙𝑒𝑠𝑠 𝑡ℎ𝑎𝑛 𝑥

• 𝑝𝑒𝑟𝑐𝑒𝑛𝑡𝑖𝑙𝑒 𝑜𝑓 𝑥 = ∗ 100
𝑡𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟𝑠 𝑜𝑓 𝑣𝑎𝑙𝑢𝑒𝑠
 EXAMPLE ON PERCENTILE
• YOU SCORED 87/100 ON A TEST.
• 39 PEOPLE SCORED LESS THAN YOU.
• THERE ARE 54 PEOPLE IN YOUR CLASS.
• WHAT PERCENTILE ARE YOU SCORING?

• ARE YOU SCORING 87 PERCENTILE? IS THAT WHAT IT MEANS?!!!!!!!!!!!!!

• WHAT MEANS PERCENTILE IS TO COMPARE YOUR DATA VALUE TO THE OTHER PEOPLE.

39
𝑝𝑒𝑟𝑐𝑒𝑛𝑡𝑖𝑙𝑒 𝑜𝑓 ′87′ = ∗ 100 = 72𝑛𝑑 𝑝𝑒𝑟𝑐𝑒𝑛𝑡𝑖𝑙𝑒 −→ P72
54
PERCENTAGE IS YOUR SCORE, BUT
PERCENTILE CALCULATES YOUR PLACE IN THE CLASS

YOU CAN GO BACK IN FOURTH IN THE EQUATION, FOR EXAMPLE IF I GIVE YOU THE PERCENTILE AND YOU FIND THE
PERCENTAGE.
IN THIS EXAMPLE YOU SCORED BETTER THAN 72 PEOPLE
 PERCENTILE
• WHAT IS ….
• P25= Q1
• P50= M
• P75= Q3
 WHAT IS IQR?
WHAT IS IQR?
• IT IS ENTERAL QUARTILE RANGE: IT IS THE DIFFERENCE BETWEEN THE QUARTILES, HOW MUCH IS THE
RANGE.
• IQR: Q3 - Q1
• MIDDLE 50% OF THE DATA

BOX PLOT
• FIVE NUMBER SUMMARY Min Max
• MINIMUM
• Q1 Q1 M Q3
• MEDIAN
• Q3
• MAXIMUM
 EXAMPLE
• 1 , 4 , 5 , 5 , 9 , 12 , 13 , 13 , 15 , 21

• MINIMUM:
• Q1:
• MEDIAN:
• Q3:
• MAXIMUM:
 EXAMPLE
• 1 , 4 , 5 , 5 , 7 , 9 , 12 , 13 , 13 , 15 , 21
Q1 M Q3

• MINIMUM: 1
• Q1: 5
• MEDIAN: 9
• Q3: 13
• MAXIMUM: 21 1 21

5 9 13
• WHAT IT MEANS AN OUTLIER? OUTLIER
• HOW WE CALCULATE THE OUTLIER? IS 21 IN PREVIOUS EXAMPLE AN OUTLIER?
• THERE IS A MATHEMATICAL WAY TO CALCULATE THE OUTLIER.
TO FIND AN OUTLIER ?
1- FIRST NEED TO FIND IQR
• IQR IN PREVIOUS EXAMPLE=Q3-Q1→ 13-5= 8
2- THE NEXT 1.5*(IQR) → 1.5*(8)= 12
• YOU WILL TAKE THE PREVIOUS NUMBER, WILL SUBTRACT IT FROM Q1, AND ADD IT TO Q3
Q1 – 1.5*(IQR)
Q3 + 1.5*(IQR)
• ANY NUMBER OUTSIDE THAT RANGE WILL BE AN OUTLIER
• 5 – 12 = -7 Q1 Q3
• 13+12 = 25 21
1
• IS THERE ANYTHING IN YOUR DATA LESS THAN -7,
OR GREATER THAN 25? 5 9 13
• NO
• THEN THERE ARE NO OUTLIERS -7 (5-12) (13+12) 25
• IF INSTEAD OF 21 I HAVE 32 THEN IT WILL BE OUTLIER
 WHAT IS Z-SCORE

• THE PERCENTAGE OF DATA FALLING WITHIN RANGE

• THE COEFFICIENT OF VARIATION
• THE NUMBER OF STANDARD DEVIATIONS AWAY FROM THE MEAN