2计量资料的统计描述仲来福7(修改)（2 statistical measurement data described in SEC 7 (modified) rifle）.docVIP

2计量资料的统计描述仲来福7(修改)（2 statistical measurement data described in SEC 7 (modified) rifle） Statistical analysis of the ninth chapter numerical variable data The first section describes the statistical numerical variable data The frequency distribution, a numerical variable data (a) establishment step frequency distribution table 1, calculate the range (range R) 2, determine the number of groups, groups and group distance. The habit is divided into 8 to 15 groups, The number of =4.45/12=0.37 group from i=R/ group (mmol/L), i=0.40 (mmol/L) will be rounded 3, a list of designated note: data were obtained in frequency statistics table 9-2 frequency column, the table is called the frequency distribution table. The frequency distribution type: normal distribution and skewed distribution (positively skewed distribution, negative skewness distribution) (two) use frequency distribution table (understand) 1. revealing the distribution characteristics of frequency (concentration and dispersion tendency) 2. types of frequency distribution revealed 3. easy to find suspicious value. 4. for further calculation of statistical indicators and statistical analysis. Two, description of central tendency (average index) Describe the common indexes: 1, arithmetic mean (arithmetic mean) 2, the geometric mean (geometric mean G) 3, the median (median) M (a) arithmetic mean (arithmetic mean) Applicable condition: used to describe the central tendency of the normal distribution data (average). 1. calculation methods of mean (1) direct method (applicable to the unit of observation is less) (2) weighting method (applicable to large samples, the frequency table data) (two) geometric mean Application: The geometric mean of 1. commonly used in geometric series data or lognormal distribution data (i.e., applicable conditions). 2. observations can not have 0. 3. observations cannot have both positive and negative values. 4. geometric same data obtained were number less than average. Calculation of geomet