Technology Considerations,Discussion:

Name of the student
Professor’s Name
Technology
5th November 2015

Discussion and Conclusion
The potential issues and limitations of my results may be briefed as follows. First of all, the statistical tests conducted provided ample evidence that the mean household income of the individuals in 2013 is significantly greater than 2012 (p<0.05). This is because a p-value less than 0.05 signifies that the difference in mean household income in between the two years has not happened due to chance factors of random sampling. Since, out of 100 observations, less than five observations of mean household income being same in both the years could be noted.

Therefore, the null hypothesis was rejected and alternate hypothesis was accepted. The alternate hypothesis indicates that since chances of the null hypothesis being too are very low, it may be inferred that there is the significant difference in mean household incomes between 2012 and 2013. Hence, the results were statistically significant. Moreover, the results were confident and could be reproduced as out of 100 observations, 95 individuals would mention that their salary is between 35364.641 to 46031.136 (95% Confidence Interval). The mean household income was 40697, which falls between these two limits. Therefore, the results are practicable too.

However, the study suffers from certain limitations. The chances of bias were pretty high since it was an interview and no financial instruments were judged to assess the actual incomes.

Further, the sampling done was an accidental one as near and dear family members or friends were involved. Such sampling may lead to assessment of income from a particular community only, where incomes may not deviate much. Therefore, the representation of data from the entire population based on age might not have been captured. Moreover, the study included income as a function of age group and not by employment or nature of work. It may happen that certain individuals were not captured in the study that was engaged in either high paid or low paid jobs, as because they did not belong to the community where research was conducted.

Further, the research may have suffered from both Type-1 and Type-11 errors. As it is known that type-1 errors indicate a failure to reject a true null hypothesis (often called false positive), while type –II error indicates a failure to reject a false null hypothesis. That means type-1 error incorporates an endpoint to be significant when there is no such significance. On the other hand type -11 errors incorporate a non-significance of results while real significance occurs. From the above limitations, it might have been possible that due to improper representation of samples, where individuals are employed in jobs, with better increment in salary on a year on year basis. This may result in Type-1 error as the study might have failed to reject a true null hypothesis.

On the other hand, it might be possible that individuals who were included in the sample did not have an improved increment structure and hence the differences that seem to be significant may be more significant. However, it can be at least portrayed that the study had reduced type-II errors, as the null hypothesis was quite truly rejected. On the other, hand the study never incorporated issues like job changes, during 2012 to 2013, either on the higher side, or on the lower side. This could have impacted the findings of the study.
Therefore, to assess whether age group is related to income and income changes as a function of time various confounding variables should be standardized. Such variables must include the proper representation of the working population from various job sectors and various socio-economic backgrounds. A stratified multistage random sampling could be done in future studies to provide power to our research based on sample demographics. Changes in the job should be taken as an exclusion criterion, if the research is conducted on the same individuals in subsequent years.