The news report did not mention the kind of statistical procedure used in the study. As mentioned in the news report, I agree that it is valid and necessary to combine land and marine temperature in order to measure the average global temperature. Basically, the temperature in land and marine areas are different from each other during day and night. The idea of combining the temperature of land and marine areas is to enable the statistician determine the average global temperature within a given time frame.
To determine whether a statistic is statistically significant, it is necessary to get the difference between a sample statistic and a specific population parameter (numerator) before dividing the value by the standard error (denominator) (Urdan, 2005, p. 64). For this reason, it is more difficult to confirm the statistical procedures as mentioned in the news report is valid given that the length of years used in each period of time varies from one another.
and vice versa. It means that when the sample size is small, bigger the standard error will be. Since a bigger sample size will have smaller standard error, there is a higher probability that the research findings will be statistically significant because of the bigger value as a result. Because of the effects of sample size in determining whether a hypothesis is statistically significant or not, Urdan (2005) explained that it is possible that a small difference between the sample statistic and population parameter to be statistically significant provided that the sample size is big (p. 64). This concept was confirmed by Phil Jones when he explained that it is more likely to achieve a statistically significance given that the available data is good for a very long period of time (Harrabin, 2009, question B).
When asked whether the global warming between 1860 – 1880, 1910 – 1940, and 1975 – 1998 were identical, the