ONE WAY ANOVA
One-way analysis of variance (abbreviated one-way ANOVA) is a technique used to compare means of two or more samples (using the F distribution). This technique can be used only for numerical data.
The ANOVA tests the null hypothesis that samples in two or more groups are drawn from populations with the same mean values. To do this, two estimates are made of the population variance. These estimates rely on various assumptions. The ANOVA produces an F-statistic, the ratio of the variance calculated among the means to the variance within the samples. If the group means are drawn from populations with the same mean values, the variance between the group means should be lower than the variance of the samples, following the central limit theorem. A higher ratio therefore implies that the samples were drawn from populations with different mean values. Descriptives | | N | Mean | Std. Deviation | Std. Error | 95% Confidence Interval for Mean | Minimum | Maximum | | | | | | Lower Bound | Upper Bound | | | QUALITY | 1 | 19 | 3.89 | .809 | .186 | 3.50 | 4.28 | 2 | 5 | | 2 | 12 | 3.83 | .937 | .271 | 3.24 | 4.43 | 1 | 5 | | Total | 31 | 3.87 | .846 | .152 | 3.56 | 4.18 | 1 | 5 | PRICE | 1 | 19 | 2.95 | .911 | .209 | 2.51 | 3.39 | 1 | 5 | | 2 | 12 | 2.75 | 1.055 | .305 | 2.08 | 3.42 | 1 | 5 | | Total | 31 | 2.87 | .957 | .172 | 2.52 | 3.22 | 1 | 5 | BRAND | 1 | 19 | 4.11 | .809 | .186 | 3.72 | 4.50 | 3 | 5 | | 2 | 12 | 4.17 | .577 | .167 | 3.80 | 4.53 | 3 | 5 | | Total | 31 | 4.13 | .718 | .129 | 3.87 | 4.39 | 3 | 5 | POST-PURCHASE SERVICE | 1 | 19 | 3.42 | .961 | .221 | 2.96 | 3.88 | 2 | 5 | | 2 | 12 | 3.67 | .778 | .225 | 3.17 | 4.16 | 3 | 5 | | Total | 31 | 3.52 | .890 | .160 | 3.19 | 3.84 | 2 | 5 | PAYMENT METHOD | 1 | 19 | 3.37 | .831 | .191 | 2.97 | 3.77 | 2 | 5 | | 2 | 12 | 2.92 | .996 | .288 | 2.28 | 3.55 | 1 | 4 | | Total | 31 | 3.19 | .910 | .163 | 2.86 | 3.53 | 1 | 5 | PACKING | 1 | 19 | 3.58 | .692 | .159 | 3.25 | 3.91 | 3 | 5 | | 2 | 12 | 3.58 | .669 | .193 | 3.16 | 4.01 | 3 | 5 | | Total | 31 | 3.58 | .672 | .121 | 3.33 | 3.83 | 3 | 5 |

Test of Homogeneity of Variances | | Levene Statistic | df1 | df2 | Sig. | QUALITY | .185 | 1 | 29 | .670 | PRICE | .590 | 1 | 29 | .449 | BRAND | 2.455 | 1 | 29 | .128 | POST-PURCHASE SERVICE | .698 | 1 | 29 | .410 | payment policy satisfaction | .176 | 1 | 29 | .678 | PACKING | .060 | 1 | 29 | .808 |

ANOVA | | Sum of Squares | df | Mean Square | F | Sig. | QUALITY | Between Groups | .028 | 1 | .028 | .037 | .848 | | Within Groups | 21.456 | 29 | .740 | | | | Total | 21.484 | 30 | | | | PRICE | Between Groups | .287 | 1 | .287 | .305 | .585 | | Within Groups | 27.197 | 29 | .938 | | | | Total | 27.484 | 30 | | | | BRAND | Between Groups | .028 | 1 | .028 | .052 | .821 | | Within Groups | 15.456 | 29 | .533 | | | | Total | 15.484 | 30 | | | | POST-PURCHASE SERVICE | Between Groups | .444 | 1 | .444 | .552 | .463 | | Within Groups | 23.298 | 29 | .803 | | | | Total | 23.742 | 30 | | | | PAYMENT METHOD | Between Groups | 1.501 | 1 | 1.501 | 1.865 | .183 | | Within Groups | 23.338 | 29 | .805 | | | | Total | 24.839 | 30 | | | | PACKING | Between Groups | .000 | 1 | .000 | .000 | .986 | | Within Groups | 13.548 | 29 | .467 | | | | Total | 13.548 | 30 | | | |

Since the level of significant of Levene Statistic for each of the items are greater than 0.05, so the results of the ANOVA table below are reliable

The table ANOVA results above shows assessment of the statistical significant of the result among items. We are going to test these variables such as QUALITY, PRICE, BRAND, POST-PURCHASE SERVICE, PAYMENT METHOD, and PACKAGING. The significant level is greater than 0.05. So we conclude that both of groups have the same idea toward those items in term of satisfaction The descriptive table results show the same level of satisfaction of two groups toward variables. Both of two clusters are neutral with QUATILITY factor (Total mean 3.87) with the range from 1 to 5 (Rate of satisfaction which are 1: Not satisfied; 2: Less satisfied; 3: Normal; 4: Satisfied; 5: Extremely satisfied). PRICE is less satisfy (Total mean 2.87) and also the range (both groups have the same range from minimum 1 to maximum 5). Moreover, PRICE is the least satisfied factor among others. On the other hand, BRAND results are significantly high satisfaction whose total mean is 4.13 and the range is narrower from 3 to 5. Lastly, both two groups of customer still remain neutral attitude toward POST-PURCHASE SERVICE, PAYMENT METHOD and PACKAGING. Their total means are 3.52; 3.19; 3.58 and their range is fluctuating from 1-5; 2-5 and 3-5.
WARD METHODS
In statistics, Ward's method is a criterion applied in hierarchical cluster analysis. Ward's minimum variance method is a special case of the objective function approach originally presented by Joe H. Ward, Jr. Ward suggested a general agglomerative hierarchical clustering procedure, where the criterion for choosing the pair of clusters to merge at each step is based on the optimal value of an objective function. This objective function could be "any function that reflects the investigator's purpose." Many of the standard clustering procedures are contained in this very general class. To illustrate the procedure, Ward used the example where the objective function is the error sum of squares, and this example is known as Ward's method or more precisely Ward's minimum variance method. YEAR WARD METHOD

Crosstab | | Ward Method | Total | | 1 | 2 | | YEARS | 1 | Count | 2 | 4 | 6 | | | % within Ward Method | 10.5% | 33.3% | 19.4% | | 2 | Count | 9 | 3 | 12 | | | % within Ward Method | 47.4% | 25.0% | 38.7% | | 3 | Count | 8 | 5 | 13 | | | % within Ward Method | 42.1% | 41.7% | 41.9% | Total | Count | 19 | 12 | 31 | | % within Ward Method | 100.0% | 100.0% | 100.0% |

Chi-Square Tests | | Value | df | Asymp. Sig. (2-sided) | Pearson Chi-Square | 2.928a | 2 | .231 | Likelihood Ratio | 2.923 | 2 | .232 | Linear-by-Linear Association | .684 | 1 | .408 | N of Valid Cases | 31 | | | a. 3 cells (50.0%) have expected count less than 5. The minimum expected count is 2.32. |

The Pearson Chi-Square result from the Chi-Square Tests table shows the reliability which is based on customer’s year of partnership. In this situation, the information is not reliable (p>0.05). According to the crosstab, the gap of the two groups customer between 3 years in considerable (YEAR 1: cooperate for less than 5 years; YEAR 2: cooperate from 5 to 10 years; YEAR 3: cooperate for more than 10 years). So there are 6 companies who have been the customer for VN Steel for less than 5 years (19.4% from total 31 customers) and also 12 customers have been partner for about 5 to 10 years and 13 customers last for more than 10 years. QUANTITY WARD METHOD

Crosstab | | Ward Method | Total | | 1 | 2 | | QUANTITY | 1 | Count | 9 | 11 | 20 | | | % within Ward Method | 47.4% | 91.7% | 64.5% | | 2 | Count | 6 | 1 | 7 | | | % within Ward Method | 31.6% | 8.3% | 22.6% | | 3 | Count | 4 | 0 | 4 | | | % within Ward Method | 21.1% | 0.0% | 12.9% | Total | Count | 19 | 12 | 31 | | % within Ward Method | 100.0% | 100.0% | 100.0% |

Chi-Square Tests | | Value | df | Asymp. Sig. (2-sided) | Pearson Chi-Square | 6.523a | 2 | .038 | Likelihood Ratio | 8.114 | 2 | .017 | Linear-by-Linear Association | 5.986 | 1 | .014 | N of Valid Cases | 31 | | | a. 4 cells (66.7%) have expected count less than 5. The minimum expected count is 1.55. |

Similar to year ward method, it seems that basing on the quantity of steel of two clusters is reliable (p<0.05). In this case, most of them only ship less than 1000 ton of steel per month (QUANTITY 1: less than 1000 tons of steel/month; QUANTITY 2: from 1000-2000 tons of steel/month; QUANTITY 3: greater than 2000 tons/month) which are 20 of them from the total 31 customers (64.5%). Seven customers ship from 1000 to 2000 ton per month (22.6%) and only 4 ship more than 2000 tons.