Problem Summary
The purpose of this presentation is to decide whether the new invention promoted by the research department should be used for the appliances in the future. The research department of an appliance manufacturing firm has developed a solid-state switch for its blender that the research department claims will reduce appliance returns under the one-year full warranty by 3%–6%. The claim needs to be tested before making the final decision about using the switch in the future.
Method Selection
The purpose of the analysis will be to determine if the proportion of returned blenders with new switches were deferent from the proportion of blenders with old switches. If the difference between the proportion is statistically significant, the new switch will be considered more desirable. In order to test for proportions, two independent samples of blenders need to be taken and the proportion of returned blenders need to be compared. The comparison can be conducted using the Z-test for differences in proportions as recommended by Zou and collogues.
Analysis Flow Chart
The present slide demonstrate a flow chart of analysis. First, null and alternative hypotheses need to be stated. Second, data needs to be collected and organized. Third, the statistical test needs to be run after testing for assumptions. And finally, the results need to be interpreted.
Hypotheses Formulation
The null hypothesis is that there is no significant difference between the proportions of returned blenders with new and old switches. The alternative hypothesis is that there is a significant difference between the proportions of returned blenders with new and old switches.
Data Collection
Data from two samples of 250 blenders was collected using simple random sampling. Simple random sampling was used because it helps to decrease the level of bias during the participant selection procedures, as everyone in the population receives an equal chance to participate.
Results
The test was performed after checking if all the assumptions were met. In particular, both samples were random and independent. The test results revealed that there was no significant difference between the proportions of returned blenders with new and old switches with a significance level of 0.05. The p-value was 0.075, which was below the significance level. Therefore, the null hypothesis was accepted.
Conclusion
The results demonstrated that the claim made by the research department was false. Since there was no difference in the number of returns, it is not recommended to decide which switch to use based on this parameter. Further research should be conducted to improve the quality of switches.
References
- Etikan, I., & Bala, K. (2017). Sampling and sampling methods. Biometrics & Biostatistics International Journal, 5(6), 00149.
- Zou, K. H., Fielding, J. R., Silverman, S. G., & Tempany, C. M. (2003). Hypothesis testing I: proportions. Radiology, 226(3), 609-613.