**7. Conclusions**

Our purpose in this paper has been to encourage authors to consider ways to improve their research. Our paper began by emphasizing the importance of careful model building, data selection, and carpentry as well as the disclosures essential to assuring the integrity and reproducibility of the work. The choice of topic should be based on its relevance to practice or on improving research techniques.

Researchers must understand the limitation of relying on *p*-values as the sole or even primary support for their findings. Confidence intervals should replace reporting *p*-values. Until the reader is informed of the importance of the results, framed by the economic or behavioral consequences of the findings, the research objective has not been achieved. The issues addressed in this paper are advanced as a means of accomplishing this objective. To assist in the mission, we strongly encourage researchers to read the paper by Greenland et al. 2016, which o ffers a tutorial on statistical tests, *p*-values, confidence intervals, and power. (Greenland et al. 2016 is also available through Dave Giles' Blog of 2 May 2019 (Giles 2019) and Bob Jensen's Blog 2 September 2019 (Jensen 2018). The Jensen blog includes the article).

**Author Contributions:** Only the categories that apply to our paper are covered here. Unless indicated, the authors contributed equally. Data curation was mainly obtained from prior research papers of the authors and papers authored by others in and outside our field. The original draft was primarily undertaken by the first author (T.R.D.) while the second author (S.A.Z.) took the lead on project review and editing. The project was under our own supervision and there was no outside project administration. The project was entirely funded and written by the authors, T.R.D. and S.A.Z.

**Funding:** The project received no outside funding.

**Conflicts of Interest:** There were no conflicts of interest.
