Stats: Hypothesis Testing Steps (Guide)

Helpfulness: 0
Set Details Share
created 6 years ago by llozada15
143 views
8 steps, names, and needed date
updated 6 years ago by llozada15
Subjects:
statistics
show moreless
Page to share:
Embed this setcancel
COPY
code changes based on your size selection
Size:
X
Show:
1

What is the __%; C.I.? (80%, 90%, 95%, 98%, 99% tipically)

.80 or tail area of .1 = z=1.28

.90 or tail area of .05 z= 1.645

.95 or tail area of .025 z= 1.96

.98, or tail area of .01 z=2.33

.99, or tail area of .005 z= 2.575

2

Step 1: State the Hypothesis

What are the null and alternative hypotheses?

  • Look at wording and sign
3

Step 2: Critical points: what is (are) the critical values?

U or P?

  • For mean u, is sigma known or unknown?
    • Known sigma; z table, need: area under tail: “guided by simple of Ha Draw a graph
      • 2 sided (a/2) (you will have +/- z) or
      • 1 sided? (You will have either a –z or +z) left or right?
    • Unknown sigma: t table: you will need: df and a or a/2 if two sided. Draw a graph
      • Use t table.
      • If right sided, you will have a +tcrit
      • If left sided, you will have a –tcrit
      • If both sided, you will have a +/-tcrit
  • For population p, is n large? Yes in this class n>30
    • Use z table Draw a graph
      • 2 sided (a/2) (you will have +/- z) or
      • 1 sided? (You will have either a –z or +z) left or right?
4

Step 3: Decision Rule: Rejection Region: reject Ho if:

  • “Z or T stat is in the region of Z or Tcrit”
5

Step 4: Find test statistic:

  • For mean U, sigma known Zstat= x-bar – U(that we want) / (sigma/sqrt(n))
  • For mean U, sigma UNknown Tstat= x-bar – U(that we want) / (st. deviation (s) /sqrt(n))
  • For population p, n>30, large = Zstat=p-bar-po(that we want)/sqrt( (po(tha we want)*(1-Po(that we want)/n)
6

Step 5: Decision:

Fail to reject Ho, or reject the Ho

  • You need zcrit and zstat or tcrit and tstat
7

Step 6: What is the conclusion?

always write it in terms of the alternative Ha

8

Step 7: What is the p value:

Unsure of how to do:

  • You do need Zstat or Tstat.
  • You also need the direction of Ha, and to know if it’s a right or a left or both sided.