Anda di halaman 1dari 7

Assignment on Research Methodology

Submitted By:

Submitted To: Prof. Dr. Uma Shankar Mishra

Management Stream Regd. No.:

IBCS, SOA University

Assignment on Research Methodology

Submitted By: Sudipta Kumar Jana Management Stream Regd. No.:

Submitted To: Prof. Dr. Uma Shankar Mishra IBCS, SOA University

Malhotra Spices Co. Pvt. Ltd.


The company was in the business of manufacturing and selling spices suitable for the Indian kitchen. The packaging of spices was in tetrapacks and the Chairman, Mr. Malhotra, wanted to know whether it could be changed to plastic or glass bottle packaging. Mr. Malhotra wanted to study the impact of different type of packaging (plastic, glass and tetrapack) on the sales of spices. Therefore, the sales are a dependent variable and type of packaging is an independent variable. Sales are a ratio scale variable, whereas the type of packaging is a nominal scales variable. If there is a difference in average sales due to various type of packaging, then it would be interesting to find out which type of packaging has the maximum impact on sales.

A one-way ANOVA is proposed to find out the impact of different packaging on sales. The hypothesis to be tested is

H0 : 1 = 2 = 3 (Average sales of different packaging is same.) H1 : All s are not same. (Average sales of different packaging are not same.) The analysis is carried out using SPSS software. The results are as under in Table 1 & 2. TABLE - 1
Des criptives Sales (in Rs. Lakhs) 95% Conf idence Interval f or Mean Low er Bound Upper Bound 122.89 141.11 97.42 111.38 111.45 125.15 112.45 124.01

N Plas tic Glass Tetra pac ks Total 10 10 10 30

Mean 132.00 104.40 118.30 118.23

Std. Deviation 12.728 9.755 9.581 15.480

Std. Error 4.025 3.085 3.030 2.826

Minimum 110 90 105 90

Max imum 150 125 130 150

Table 1 indicates that the mean sales due to plastic packaging are the highest whereas due to glass packaging, it is the lowest. The standard deviation in sales due to plastic packaging is large as compared to the other two types of packaging where it is almost equal.

To test the hypothesis stated above, results of one-way ANOVA are presented in Table 2 below: TABLE - 2
ANOVA Sales (in Rs. Lakhs) Sum of Squares 3808.867 3140.500 6949.367 df 2 27 29 Mean Square 1904.433 116.315 F 16.373 Sig. .000

Betw een Groups Within Groups Total

The results indicate that computed

. Assuming a 5% level of significance, the table F value is

. Since computed F is greater than tabulated F, null hypothesis is rejected. Therefore average sales is different due to different type of packaging.

To know which type of packaging has the highest sales and which one has the lowest sales, we test the following hypotheses:

H0 : 1 = 2 (sales of plastic and glass packaging is same.) H0 : 1 > 2 (sales of plastic packaging is higher than that of glass packaging.)

A t test for difference between two means is conducted and results are presented in Table 3 & 4.

TABLE - 3
Group Statis tics Ty pe of Packaging Plas tic Glass N 10 10 Mean 132.00 104.40 Std. Deviation 12.728 9.755 Std. Error Mean 4.025 3.085

Sales (in Rs. Lakhs)

TABLE - 4
Inde pe nde nt Sam ples Te st Levene's Test f or Equality of V ariances t-test f or Equality of Means 95% Conf idence Interval of the Dif f erence Low er Upper 16.946 16.894 38.254 38.306

F Sales (in Rs. Lakhs) Equal variances as sumed Equal variances not assumed 1.277

Sig. .273

t 5.443 5.443

df 18 16.861

Sig. (2-tailed) .000 .000

Mean Dif f erence 27.600 27.600

Std. Error Dif f erence 5.071 5.071

We assume 5% level of significance. The p value (zero) for both equal and unequal variance cases are less than level of significance, the null hypothesis is rejected. Therefore, the sales of plastic packaging is higher than that of glass packaging.

II

H0 : 1 = 3 (sales of plastic and tetra packaging is same.) H0 : 1 > 3 (sales of plastic packaging is higher than that of tetra packaging.)

The results of t test for the difference between two means is conducted and the results are presented in Tables 5 & 6.

TABLE - 5
Group Statis tics Ty pe of Packaging Plas tic Tetra pac ks N 10 10 Mean 132.00 118.30 Std. Deviation 12.728 9.581 Std. Error Mean 4.025 3.030

Sales (in Rs. Lakhs)

TABLE - 6
Inde pe nde nt Sam ples Te st Levene's Test f or Equality of V ariances t-test f or Equality of Means 95% Conf idence Interval of the Dif f erence Low er Upper 3.116 3.058 24.284 24.342

F Sales (in Rs. Lakhs) Equal variances as sumed Equal variances not assumed 1.041

Sig. .321

t 2.719 2.719

df 18 16.720

Sig. (2-tailed) .014 .015

Mean Dif f erence 13.700 13.700

Std. Error Dif f erence 5.038 5.038

A 5% level of significance is assumed. Since it is one tailed test, the p value corresponding to equal and unequal variances are and respectively. In both the cases p values are less than ,

the level of significance, the null hypothesis is rejected. Therefore the average sales of plastic packaging is higher than that of tetra packaging.

III

H0 : 2 = 3 (sales of glass and tetra packaging is same.) H0 : 2 < 2 (sales of glass packaging is less than that of tetra packaging.0

The results of t test for two independent samples are reported in Tables 7 & 8.

TABLE - 7
Group Statis tics Ty pe of Packaging Glass Tetra pac ks N 10 10 Mean 104.40 118.30 Std. Deviation 9.755 9.581 Std. Error Mean 3.085 3.030

Sales (in Rs. Lakhs)

TABLE - 8
Inde pe nde nt Sam ples Te st Levene's Test f or Equality of V ariances t-test f or Equality of Means 95% Conf idence Interval of the Dif f erence Low er Upper -22.984 -22.984 -4.816 -4.816

F Sales (in Rs. Lakhs) Equal variances as sumed Equal variances not assumed .068

Sig. .797

t -3.215 -3.215

df 18 17.994

Sig. (2-tailed) .005 .005

Mean Dif f erence -13.900 -13.900

Std. Error Dif f erence 4.324 4.324

Again a 5% level of significance is assumed. As it is a one-tailed test, the p value corresponding to equal and unequal variance is and 0.0025 respectively. In both the cases, the p values are less than level

of significance, the null hypothesis is rejected. Therefore average sales of tetra pack are higher than glass.

By summing up the results of three cases, we can infer that the average sales of plastic is statistically higher than tetrapack, which is statistically higher than glass.

2.

Now, treating store size as the block, there would be two independent variables. Therefore, a two-way ANOVA could be conducted. The two hypotheses to be tested are:

Packaging

H0 : 1 = 2 = 3 (Average sales due to different sizes of packaging is same.) H1 : At least two means are not same. II Stores H0 : 1 = 2 = 3 (Average sales in different types of stores is same.) H1 : At least two means are different. The results of two-way ANOVA are presented in Tables 9. TABLE - 9
Tes ts of Be tw ee n-Subje cts Effe cts Dependent V ariable: Sales (in Rs . Lakhs ) Sourc e Correc ted Model Intercept packaging store Error Total Correc ted Total Ty pe III Sum of Squares 3921.122a 413478.828 3808.867 112.256 3028.244 426323.000 6949.367 df 4 1 2 2 25 30 29 Mean Square 980.281 413478.828 1904.433 56.128 121.130 F 8.093 3413.519 15.722 .463 Sig. .000 .000 .000 .634

a. R Squared = .564 (A djusted R Squared = .495)

The computed F statistic corresponding to packaging and stores are respectively. The table value of F assuming 5% level of significance in two cases is

and . Since

computed F for packaging is higher than tabulated F, null hypothesis is rejected for packaging. Therefore, there is a difference in sales due to packaging. However, for stores, the computed F value is less than the tabulated F value, therefore the null hypothesis due to stores is accepted. This means that there is no effect of the store size on sales. This implies that average sales of different types of stores are same. 6

In this case, it is seen that the sales is influenced by the type of packaging. The size of the store has no influence on sales. Further the average sales of plastic packaging are the highest, followed by tetra packaging and glass. Therefore, it is recommended that the company should introduce plastic packaging, provided there are no technological or economic constraints.

Anda mungkin juga menyukai