Wednesday 2 May 2018

Quantitative Data Analysis


Quantitative Data Analysis

Analysis of quantitative data was used to test the hypothesis with the statistical test, but before tested statistically, then the data obtained has to go through several stages, among others:

1. 

Increase your raw data into raw data (data from data transformation ordinal data into intervals with these steps:
  • Find the largest and smallest score · Look for the value range 
  • Looking for a large number of classes
  • Looking for the value of the length of the class 
  • Create a table with tabulated helper 
  • Find the average (mean) · Looking for raw Byway 

Changing the ordinal data into data interval by the formula:

T 1 = 50 + 10 (Xi – x) S 

2.

Testing of its homogeneity through Bartlet with test steps:
  1. Enter the figures of statistics for testing its homogeneity on the chart helper · Calculates the variance of a sample third · Calculate Log S · Calculate the value of B
  2. Calculate the value of X ² count = (lon 10) (B – Ʃ (dk) Log Si²)
  3. Compare the value X ² calculate with X ² table with rules If X ² count ≥ X ² chart, meaning is not homogeneous and If X ² count ≤ X ² tables, means homogeneous

3. 

Testing the normality of data by the method of chi squared with these steps:
  • Find the largest and smallest score 
  • Look for the value range 
  • Looking for a large number of classes 
  • Looking for the value of the length of the class 
  • Create a table with tabulated helper 
  • Find the average (mean) 
  • Looking for raw Byway 
  • Make a list of the expected frequency by means of: 
  • Determine the limits of the class 
  • Find the value of Z – score 
  • Looking for a spacious o – z 
  • Looking for breadth of each class interval 
  • Find the expected frequencies (fe) 
  • Looking for a chi squared count 
  • Compare the value X ² calculate with X ² table with rules
If X ² count ≥ X ² chart, This means that the data distribution is not normal and If X ² count ≤ X ² tables, meaning that the normal distribution of the data

4. 

Literariness regression testing with these steps:
  • Looking for statistics; X, Y, X², Y², XY, x, a, b 
  • Find the amount of squares regression (JKReg (a)) 
  • Find the amount of squares regression (JKReg (bja)) 
  • Look for the number of quadratic residues (JKRes) 
  • Find the average number of squares regression (RJKReg (a)) 
  • Find the average number of squares regression (RJKReg (bja)) 
  • Find the average number of quadratic residues (RJKRes) 
  • Find the amount of squares error (JKE) · Find the amount of squares of tuna is suitable (JKTC) 
  • Find the average number of squares of tuna (RJKTC) 
  • Find the average quadratic error (RJKE) 
  • Find the value of F count 
  • Find the value of F table 
Compare the value F count with F table with rules If F count ≥ F tables, meaning that the data is not linear and patterned If F count ≤ F tables, meaning that data is linear Next to hypothesis testing with use the statistics k simple regression, regression, correlation and simple correlation.

5. 

Simple regression equations Formulas:

Ŷ = a + b 1 6.
The regression equation Formula doubles:
Ŷ = a + b 1 X 1 + b 2 X 2 7.
Pearson Product Moment Correlation formula (PPM):
n (ƩXY)-(ƩX). (ƩY) 1 r count = {n. ²-ƩX (ƩX) ²}. 

{n. ²-ƩY (ƩY) ²} 
where: r count = koefesien correlation ƩXi = total score items ƩYi = total score total (all items) n = number of respondents Correlation of PPM is denoted (r) with the provisions of the r value is no more than the price (-1 ≤ r ≤ + 1).
When:
r =-1 means perfect negative correlation r = 0 means that there is no correlation r = 1 meaning very strong correlation,

While the meaning of the price r will be consulted with tables of the interpretation value of r as follows:

1. 0.80 – 1.000 = very strong 2. 0.60 – 0.799 = Strong 3. 0.40 – 0.599 = strong enough 4.
0.20 – 0.399 = Low 5. 0.00 – 0.199 = very low The next big small donations to declare variable X against Y can be determined with the formula determinant coefficient as follows:

KP = r ² x 100% where: KP = value determinant coefficient

R = the value of the coefficient correlation Advanced testing i.e. test the significance of that Act if want to find the meaning of relationship variable X against Y, the results of the tested with PPM correlation test of significance of the formula:

 r n-2 t count = 1 – r where: t count = the value of t r = the value of the coefficient correlation n = number of samples Distribution (table t) for α = 0.05 and degrees of freedom (dk = n – 2) Decision Rule:

If t count > t table means significant If t count < t table means insignificant Multiple correlation Analysis to test the hypothesis whether or not there are significant effects simultaneously between capability (X1) and work motivation (X2) against the performance of employees (Y) using the formula:

 r²x 1. y + r²x 2. y -2 (rx 1. y ). (rx 2. y ). (rx 1. X 2) The X 1. X 2. 

The = 1- the ² X 1. X 2 Next to know the significance of a correlation first sought double F count later compared to F table .

R² F count = k (1- R²) n- k-1 

where:

R = correlation coefficient double value 
k = number of free variables (the independent) 
n = number of samples 

Significance of testing rules:

If F count ≥ F table , then reject Ho means significant F count ≤ F table thanks Ho meaning not significant Look for the value F table using a table of F with the formula:

Significant Extent: α = 0.01 or α = 0.05 F = table {F(1 – α) (dk = k), (dk = n-k-1)}

No comments:

Post a Comment