In clinical and experimental studies is quite common to use a small sample, when the number of observations is less than 30. When the sampling means and indicators are calculated using the same formulas as for the large. In the calculation of the standard deviation and the mean error indicator, the number of observations is reduced by one:
The reliability of the results (t) measured at the Table 5.12. Student contact with the table should be at column 1, and stating the number of degrees of freedom (n) equal to n-1, ie the number of observations minus one. These 2, 3 and 4th Earl calculated the probability of a correct conclusion, equal to 95% – Column 2, at risk: the errors of 5% (P05), 99% – Column 3, with the risk of error of 1% (P01) and 99 , 9% column 4, at the risk of error of 0.1% (P001).
Methods of measurement between events
One of the important tasks of the research is to identify and measure the relationship between the features characterizing the phenomena or processes. Distinguish between functional and correlation of communication.
When a change in the functional relationship of one trait will inevitably cause very definite change of the other sign. An example of such a relationship can serve as a dependent area of a circle from its radius. The functional connection between the phenomena inherent in inanimate nature. In the biological sciences often have to deal with a relationship between events when one and the same value of one attribute corresponds to a number of varying values of another trait that is caused by the interaction of various extremely diverse phenomena of nature. This kind of relationship is called correlation (correlation – a match, the correlation). At the time, as there is a functional connection in each observation, correlation occurs only with numerous comparison features.
Consider, for example, the relationship between the age of pre-school children and their growth. The data show that with age, the growth of children increases, and so we can assume there was a link between these traits.
Age 3 years 4 years 5 years 6 years 7 years
Height in cm 100.3 102.9 108.1 113.7 118.3
92.6 100.1 106.8 113.8 119.2
93.8 101.6 107.8 113.3 119.4
93.7 98.4 104.6 111.8 116.1
94.2 99.4 107.4 112.1
However, it should be noted that the same correspond to different age children’s growth. This is because the growth of children is determined not only by age, it is influenced by many other sectors, including living conditions, nutrition, physical education, etc. Thus, we can conclude that the relationship between age and growth of children is a correlation.
The researcher should be remembered that the discovery of the correlation between the compared phenomena does not say more about the existence of a causal link between the two. To establish the latter requires a comprehensive logical and ad hoc processes being studied. The statistical method allows to prove the resulting research findings about the existence of other ties between the phenomena to identify the most important of them.
The strength of the connections between phenomena, its tightness and orientation determined by the value of the correlation coefficient, which ranges from 0 to 1. When r = 0 there is no connection, for r = 1 – the connection is complete, functional.