Professional Documents
Culture Documents
29 30 31 32 33 34 35 36 37 38 39 40
22 23 19 21 20 21 21 24 25 26 23 24
211 176 160 150 165 140 151 193 178 135 187 179
1846 1540 1400 1313 1444 1225 1321 1689 1558 1181 1636 1566
x
x!
i !1
6825 ! 166.15 40
y
y!
i !1
W=
x
! 17,51
2
This result tells us that the between the real and estimated number of worked hours and the amount of salary there is a difference of , plus/minus, 17,51 hours b. The standard deviation for the amount of the vendors monthly net salary
W=
y
! 150,52
2
This result tells us that the between the estimated and the real amount of monthly salary and the number of worked hours there is a difference of , plus/minus, 150,52 hours 3. a. The coefficient of variance of the worked hours Cv ! W 100 ! 101,68% x
Cv !
W 100 ! 100,35% y
Because the level of the coefficient of variance is over the level of 35% we can conclude that the average worked hours and the average monthly net salary is not representative and the data is not heterogeneous. Table 2: The Average, Standard Error and Standard Deviation
Frequency distribution
The 40 vendors will be grouped in 5 intervals with the calculated range of 15.2 for the variable x and of 133 for the variable y.
7 6 16 9 2
7 7 17 7 2
x f n
i i
! 170,34
b. The arithmetic mean of the amount of the monthly net salary of the n=40 vendors: y!
y f n
i i
! 3289,44
This result tells us that between the real and estimated number of worked hours and the amount of salary there is a difference of, plus/minus, 13, 05 hours. b. The standard deviation for variable y:
2 W y ! W y ! 57,35
This result tells us that the between the estimated and the real amount of monthly salary and the number of worked hours there is a difference of , plus/minus, 57,35 hours 3. a. For variable x, the number of worked hours the coefficient of variance is : Cv ! W x 100 ! 7,66%
b. For the y variable, for the amount of the monthly net salary the coefficient of variance is: Cv ! W y 100 ! 1,74%
Because the coefficient of variance of the two variables is below 35% it results that the grouping of x and y is eloquent. Table 3: The frequency distributions
18 16 14 Frequency 12 10 8 6 4 2 0
135-135
150.2-150.2
180.6-180.6
195.8-195.8
20
15 Frequency
10
This charts give us useful information about the shape of the distribution and as we can see above for both of the variables the highest number of data is found the third interval.
Salary frequency 33 39
% 82.5 97.5
y s 2W
68% of the data will fall within 1 standard deviation of the mean 95% of the data will fall within 2 standard deviations of the mean Almost all (99.7%) of the data will fall within 3 standard deviations of the mean
Regression
Relationship Betweem Worked Hours And The Monthly Net Salary
2000 1500 Salary 1000 500 0 0 50 100 Worked Hours 150 200 250 y = 8.3514x + 57.868 R = 0.9435
From this scatter diagram we can see that the model is almost a perfect line, a linear model. This means that the two variables are interdependent and for example the amount of the monthly net salary depends on the number of worked hours in the company.
Because Multiple R = 0,97 we can draw the conclusion that the link between the amount of the vendors monthly salary and the number of hours they work are linked together. The Standard Error reflects the difference between the estimations made about that there is an average difference of 36,25 between the two chosen variables.
ANOVA Regression Residual Total SS MS F Significance F 1 833697.9974 833697.9974 634.3237 2.58866E-25 38 49943.77762 1314.309937 39 883641.775 Standard Coefficients t Stat P-value Error 57.868 56.86746539 1.017593443 0.315307 8.3514 0.331591679 25.18578452 2.59E-25 df
From the table above, we can see that the intercept variable is 57,868 and it mean that when the explanatory level is 0, when the number of worked hours is 0, the amount of the monthly net salary will be of 57,868 lei. Because the value of P-value is high 0,31 the coefficient is insignificant. The coefficient b = 8,35 and it means that if the level of worked hours will increase also the salary will increase with 8,35 lei.
I Coefficient of correlation test. 1. Formulation of hypotheses: H0: r = 0 H1: r { 0 2. We choose a significance level of 95% ( = 0.05); 3. Because the number of observations (40) is higher than 30 we will choose Student test (z) with n-2 degrees of freedom. 4. z tab ! z E / 2;n 2 ! z 0, 025; 25 ! 1,96 n2 38 ! 0.97 ! 24,41 2 0.06 1 r
z calc ! r 5.
accept the alternative hypothesis, therefore the model is valid II. Testing the significance of the a parameter:
1. Formulation of hypotheses: H0: = 0 H1: {0
2. We choose a significance level of 95% ( = 0.05); 3. Because the number of observations is higher than 30 we will choose Student test (z) with n-2 degrees of freedom ( in our case 38 d.f)
5.
Sa !
SSE n2
(x
i
x)2
7. Since z calc > z tab , for a significance level of accept alternative hypothesis. III. Testing the significance of the b parameter: 1. Formulation of hypotheses: H0: = 0 H1: { 0
3. Because the number of observations is higher than 30 we will choose Student test (z) with n-2 degrees of freedom ( in our case 38 d.f); 4. z tab ! z E / 2;n 2 ! z 0, 025; 25 ! 2,38 5. z calc ! b ! 27,93 ( extracted from ANOVA z Stat) Sb
Sb !
SSE n2
(x
6.
i
! 0.299
2
x)
7. Since z calc > z tab , for a significance level of accept alternative hypothesis.