red consecutively as well. For example, a paper with two graphs, a reproduction of a segment of chart record and two tables will have figures 1, 2, and 3, and tables 1 and 2. Do note that I distinguished graphs from chart records. Not everything with gridlines is a graph. Graphs are analytical tools. Chart records are raw data (which may be presented in results as an example, if appropriate).

       Do not draw conclusions in the results section. Reserve data interpretation for the discussion.

       The significance of 'significance'

       We have a statistically significant difference when analysis yields a very low probability that the difference was due to sampling error (random error) alone. If sufficient data are collected, and statistical significance is not achieved, the investigator can conclude that the null hypothesis is supported ñ there is no significant difference.

       Lack of a significant difference does not mean that the result itself is insignificant. A finding, for example, that there are no intrinsic differences in fundamental mathematical ability among racial groups would be a very significant finding. Significance in this study refers to the importance of the result. "It is significant that we found no significant differences among the groups studied" is a valid, though perhaps confusing, statement.

       There is a tendency among students to reject a study as inconclusive just because no statistically significant differences were found. Such rejection suggests a misunderstanding of the scientific method itself. You can conclude something from even the most poorly designed experiments. In fact, most well-designed experiments result in support for the null hypothesis. Be prepared to interpret whatever you find, regardless of what you think you should find. The purpose of experimental science is to discover the truth - not to make the data conform to one's expectations.

       来源：https://www.ruf.rice.edu/~bioslabs/tools/report/reporterror.html

 

上一页  [1] [2] [3] [4] [5]