Paper Summary:

The paper looks at the claim that the null hypothesis is always false, as well as the problems with misinterpretations of the significance level. The writer brings to question the editors of psychological journals for the emphasis placed on achieving a low significance level as an unofficial criterion for publishing an article. The paper attempts to informally critique the Significance Test, an alternative to the traditional NHST.

From the Paper:

"The Null Hypothesis Significance Test (NHST) exists amid a great deal of tradition and controversy. Like any theory, NHST has its strong points and its weak points. While a number of learned scholars appear to agree that the NHST has become ritualized and perhaps misinterpreted, no one seems to honestly feel that the NHST should be completely abolished (Cohan, 1994; Bakan, 1967; Rozeboom, 1960). R.A. Fisher (1951) pioneered the notion of the null hypothesis (H0), which was designed to compliment the hypothesis in which the researcher was interested. Thus the null hypothesis, in Fisher's view, was the hypothesis to be revealed as false, or "nullified" (Bakan, 1967). An example of this would be if a researcher hypothesized that Psychology students have different IQ scores than English students. The null hypothesis (H0) would state that Psychology students and English students IQ scores are equal. Therefore, rejecting the null hypothesis would logically imply that the hypothesis of interest, the alternate hypothesis (H1), was correct. The significance level (p < .05, etc.), is designed to be a measure of the probability that, given H0 is true, these (or possibly more extreme data) could occur (Cohen, 1994)."

Visitors who viewed this Essay also liked these:

• Essay # 84120 : "'The Real World of Technology'"
• Book Review # 32095 : "Ursula Franklin's "The Real World of Technology"
• Book Review # 113494 : "House and Street: The Domestic World of Servants"

APA Citation:

MLA Citation: