ONE SAMPLE HYPOTHESIS TESTING

IS n<30?

NO
YES

Z TEST

CONFIDENCE INTERVALS IS s Z CRITICAL FROM Z

(POPULATION SD) TABLES
KNOWN? x - µH0
TS =
SD
n
IS n<30?

YES
NO
N T TEST Z TEST

S
O

YE
T CRITICAL FROM T Z CRITICAL FROM Z
TABLES WITH n - 1 TABLES
DEGREES OF FREEDOM
Z TEST x - µH0 x - µH0
IS s UCL = x + za
SD TS =
SD
TS =
SD
n
(POPULATION SD) n
2 n
KNOWN?
SD TWO SAMPLE HYPOTHESIS TESTING
LCL = x - za
2 n IS AT LEAST
ONE OF
n1 , n2 < 30 ?
YE
NO

S
NO

YE
Z TEST
T TEST Z TEST ARE s 1 & s 2
(POPULATION Z CRITICAL FROM Z
SD's) KNOWN? TABLES
(x1 - x2 ) - (µ1 - µ 2 )H
s s TS = 0

UCL = x + tan -1 UCL = x + za SD12 SD22

+
2 n 2 n Y n1 n2

NO
ES

n -1 s s
LCL = x - ta LCL = x - za
2 n 2 n Z TEST
ASSUMTION
s1 = s 2 Z CRITICAL FROM Z
TABLES
(POPULATION SD's)
ARE EQUAL? (x1 - x2 ) - (µ1 - µ 2 )H
TS = 0

MARGIN OF ERROR SD12 SD22

+
n1 n2

s1
s

= ES

N s2
2

¹
O
SD Y s
1
E = za T TEST T TEST
2 n T CRITICAL FROM T TABLES T CRITICAL FROM T TABLES
2 WITH WITH DEGREES OF FREEDOM
æ SD ö n +n -2
n = ç za ÷ n1 - 1 2
DEGREES
x - µ OF FREEDOM æ s12 s22 ö
2

çç n + n ÷÷
è 2 E ø
H
TS = 0

SD è 1 2 ø
n
æ æ s 2 ö2 ö æ æ s 2 ö2 ö
çç 1 ÷ ÷ çç 2 ÷ ÷
E n (x1 - x2 ) - (µ1 - µ 2 )H ç çè n1 ÷ø ÷ ç çè n2 ÷ø ÷
za = TS = 0
ç ÷+ç ÷
æ 1 1 öæ (n1 - 1)s12 + (n2 - 1)s22 ö÷ ç n1 - 1 ÷ ç n2 - 1 ÷
2 SD ç
ç n
è 1
+ ֍
n2 ø÷çè n1 + n2 - 2 ÷ ç ÷ ç ÷
ø è ø è ø
E n (x1 - x2 ) - (µ1 - µ 2 )H
SD = TS = 0

za s12 s22
+
2 n1 n2

SCENARIO 1 H 0 :µ =
H1 : µ ¹ SCENARIO 3
Two-Tail Hypothesis Test H0 : µ ³
H1 : µ <
One-Tail Hypothesis Test
Reject Ho Reject Ho

Do Not Reject Ho

Reject Ho

Do Not Reject Ho
- za za
2 2

- tadf tadf
2 2

SCENARIO 2 H 0 :µ £
- za
H1 : µ >
One-Tail Hypothesis Test - tadf

Reject Ho

Do Not Reject Ho

za
tadf
 Summarizing the z-scores for Various Levels of a.
Alpha a Tail Critical z-score

0.01 One ± 2.330

0.01 Two ± 2.575
0.02 One ± 2.055
0.02 Two ± 2.330
0.05 One ± 1.645
0.05 Two ± 1.960
0.10 One ± 1.280
0.10 Two ± 1.645

P- Value
This is the smallest value of significance at
which the null hypothesis is rejected.

One-tail test, p-Value

For a one-tail test, with rejection region to the
right, the p-value will be the area to the right
of the test statistic. E.g. if the test statistic is
1.78 the p-value will be 3.75%.

Reject Ho
Do NOT reject Ho

3.75%

z
0 1.78
0.4625 '0.5-0.4625=0.0375

Two-tail test, p-Value

Using above example- remember to multiply area by
2, the answer will be 7.5%.
If the p-value #a, we will reject the null hypothesis

If the p-value >a, we will not reject the null hypothesis