Statistical Enquiry Cycle Conclusion Evaluation The Conclusion stage

Statistical Enquiry Cycle Conclusion / Evaluation

The Conclusion stage requires you to… 1. Answer your investigative question 2. Make a correct inference 3. Discuss the impact of sampling variability NOTE: all statements must be • justified (using numerical statistics), and • clearly linked to the context • must relate only to the population medians!

1. Answer your investigative question Conclusion 1. Answer your original investigative question • Repeat it if necessary 2. It must be consistent (correct) with the analysis (graphs, summary statistics and informal confidence intervals) 3. It will involve making a call about the population medians 4. Answer your prediction – comment on this

2. Make an inference about the population medians Conclusion Use the overlap/non-overlap to make the appropriate inference NO OVERLAP “Due to no overlap between the informal confidence intervals of males and females, I can make the inference that there is sufficient evidence to say that the population median for male heights is greater than the population median for female heights” OVERLAP “As there is overlap between the informal confidence intervals of males and females, I can make the inference that there is insufficient evidence to say that the population median for male heights is greater than the population median for female heights”

3. Discuss the impact of sampling variability Conclusion Our analysis and conclusion is based on only one set of sample comparisons. If another sample was taken, the resulting statistics, analysis, confidence intervals etc. would be different. In the majority of cases, the conclusion would stay the same (especially if the sample size >30, the sampling method was unbiased, and the sample was representative of the population). However, there is a chance that the conclusion could change (especially if you have slight overlapping, or almost overlapping samples)

The Evaluation stage requires you to… Evaluation 1. Discussion of possible errors, bias, omissions, improvements, further research, and other explanations • Was the sampling method you chose the most appropriate? If not, why? What other method would provide you with a more representative sample of the population? • Are there any biases involved? Are there any other external factors not considered that could of influenced the result? • What improvements could you make to the process, and why? • Where could this investigation take you now?

The Evaluation stage requires you to… Evaluation 2. Any further reflections on the process • Is there any other contextual insight or reflective thought you could add?