9.4
P-Wert
Der P-Wert ist eines der wichtigsten Konzepte in der Statistik.
P-Wert steht für den Wahrscheinlichkeitswert. Der P-Wert ist die Wahrscheinlichkeit, d…
Wenn eine Teststatistik aus der Stichprobenstatistik berechnet wird, z. B. der Stichprobenanteil, kann sie sich in einer Wahrscheinlichkeitsverteilung befinden.
Dieser Wert der Teststatistik grenzt einen Bereich unter der Kurve vom Rest des Bereichs ab.
Dieser Bereich am Ende der Verteilung ist der P-Wert, wobei P für Wahrscheinlichkeit steht.
Unter der Annahme, dass die Nullhypothese wahr ist, besteht immer die Möglichkeit, die berechnete Teststatistik oder einen höheren Wert im kritischen Bereich der gegebenen Verteilung zu beobachten.
Ein p-Wert gibt die Wahrscheinlichkeit an, dass dieser Teststatistikwert im kritischen Bereich allein durch Zufall erreicht wird.
Wenn also festgestellt wird, dass der P-Wert kleiner als ein vorgegebener Wert ist – z. B. 0,05 –, lehnen wir die Nullhypothese ab, da sie darauf hinweist, dass das beobachtete Ergebnis höchst unwahrscheinlich ist und die Beweise gegen die Nullhypothese stärker sind.
Der P-Wert kann je nach Hypothese oder Wert der Teststatistik am rechten Ende, am linken Ende oder an beiden Schwänzen berechnet werden.
View the full transcript and gain access to JoVE Core videos
Q1: What does P-value represent in hypothesis testing?
The P-value, or probability value, represents the probability of observing a test statistic as extreme or more extreme than the calculated value, assuming the null hypothesis is true. It measures the likelihood of obtaining sample results by chance alone. A smaller P-value indicates stronger evidence against the null hypothesis.
Q2: How do you use P-value to make decisions about the null hypothesis?
Compare the P-value to a predetermined significance level, typically 0.05. If the P-value is less than 0.05, reject the null hypothesis, indicating the observed outcome is highly unlikely under the null hypothesis. If the P-value is greater than 0.05, fail to reject the null hypothesis.
Q3: Where is the P-value located in a probability distribution?
The P-value is the area at the tail of the probability distribution, demarcated by the test statistic value. It can be calculated at the right tail, left tail, or both tails depending on the hypothesis and test statistic value. This tail area represents the probability of observing extreme results by chance.
Q4: What does a large P-value indicate about your hypothesis?
A large P-value indicates insufficient evidence to reject the null hypothesis. However, a large P-value does not mean the null hypothesis is true. It simply suggests the observed data is reasonably likely under the null hypothesis, so you fail to reject it rather than accept it.
Q5: Is P-value the same as the probability of rejecting the null hypothesis?
No. P-value is not the probability of rejecting the null hypothesis, nor is it a statistical error rate or sampling error. P-value also does not indicate a 95% chance that the observed difference is real. It specifically measures the probability of observing the test statistic under the assumption that the null hypothesis is true.
Q6: Why is 0.05 commonly used as the significance level for P-values?
The 0.05 significance level is a predetermined threshold convention in statistics. When P-value is less than 0.05, results are considered statistically significant, meaning the observed outcome is highly unlikely by chance alone. This threshold balances the need to detect real effects while controlling false positives.
Q7: What information does P-value NOT provide about hypotheses?
P-value does not convey information about the truth of the null or alternative hypotheses. It does not indicate whether the null hypothesis is true, nor does it measure the magnitude or practical importance of an effect. P-value is solely a measure of evidence strength against the null hypothesis based on sample data.