In: Statistics and Probability
Is it possible to not reject both the null hypothesis and
alternative hypothesis? It kinda sounds impossible by common sense,
so I wonder if there's an error in my working, thanks for your
help/
In a condition of n=84, (sample mean) = 0.4302, s (sample
STDV) = 0.260, d.f.= 83, level of significance 0.05,
confidence level 95%, critical value (one tailed)
1.663
I have made a null hypothesis and alternative of H₀: μ (the
"estimated" population mean) ≤ 0.453 and H₁: μ >
0.453
and by finding the computed value of t, (− μ) ÷ / =-0.8, I
concluded not to reject H₀
But what baffled me was when I repeated the same procedure, but
now the alternative hypothesis as the new null hypothesis,
thus, H₀: μ ≥ 0.453, H₁: μ< 0, I also could not reject
the hypothesis. I know can mathematically understand this because
we're using just the same numbers with a different direction of
sign of inequality, but I can't understand how both hypotheses can
be accepted.
I'll be waiting for any expert's kind replies, thank you for reading.
The choice of Null Hypothesis and Alternative Hypothesis purely depends on the research question on hand. They cannot be just interchanged at will as described in Problem 1 and Problem 2 below:
Problem 1:
In this particular investigation:
The research question is as follows:
Is the population mean greater than 0.453.
For testing the research question:
H0: Null Hypothesis: 0.453 (the population mean is not greater than 0.453.)
HA: Alternative Hypothesis: 0.453 (the population mean is greater than 0.453.) (Claim)
n = 84
= 0.4302
s = 0.260
=0.05
df = 84 -1 = 83
From Table, critical value of t = 1.663
Test Statistic is given by:
Since calculated value of t = - 0.804 is less than criticalvalue of t = 1.663, the difference is not significant. Fail to reject null hypothesis.
Conclusion:
The data do not suport the claim that the population mean
is greater than 0.453.
Problem 2:
In this particular investigation:
The research question is as follows:
Is the population mean less than 0.453.
For testing the research question:
H0: Null Hypothesis: 0.453 (the population mean is not less than 0.453.)
HA: Alternative Hypothesis: 0.453 (the population mean is less than 0.453.) (Claim)
n = 84
= 0.4302
s = 0.260
=0.05
df = 84 -1 = 83
From Table, critical value of t = - 1.663
Test Statistic is given by:
Since calculated value of t = - 0.804 is greater than criticalvalue of t = - 1.663, the difference is not significant. Fail to reject null hypothesis.
Conclusion:
The data do not suport the claim that the population mean is less
than 0.453.
Thus, we note the following:
The choice of Null Hypothesis and Alternative Hypothesis purely depends on the research question on hand. They cannot be just interchanged at will.
In this particular investigation:
If the research question is:
Is the population mean greater than 0.453.
The conclusion is:
The data do not suport the claim that the population mean is greater than 0.453.
On the conrary:
If the research question is:
Is the population mean less than 0.453.
The conclusion is:
The data do not suport the claim that the population mean is less than 0.453.