登录查看更多内容

How can you ensure study results are valid and reliable?

由人工智能和领英社区提供技术支持

Biotechnology is a field that relies on rigorous and reliable research to advance scientific knowledge and develop innovative solutions. However, conducting biostatistical analyses and reporting study results can pose ethical and regulatory challenges that require careful attention and quality control. In this article, you will learn some tips and best practices to ensure your study results are valid and reliable, and avoid common pitfalls and errors that could compromise your credibility and reputation.

此文章中的业界达人

由社区从 6 条内容中精选。了解更多

Darko Medin

Data Scientist and a Biostatistician. Developer of ML/AI models. Researcher in the fields of Biology and Clinical…
Nouran Hamza

Owner |Biostatistics | Public Health | Clinical Research | Host biostat done 4u podcast| Data to patient care decision.
Medha Sharma

Research Associate in Genomics and Bioinformatics at Celebal Technologies | Data Scientist | Deep Learning…

1 Define your research question

The first step to ensure valid and reliable study results is to define your research question clearly and precisely. Your research question should be specific, relevant, feasible, and answerable with the data and methods you have available. It should also align with the objectives and hypotheses of your study, and guide your design, sampling, measurement, and analysis choices. A well-defined research question will help you avoid unnecessary or inappropriate data collection, analysis, or interpretation, and ensure your results are focused and meaningful.

添加您的观点

2 Choose your design and methods

The next step is to choose your study design and methods that are suitable for your research question, data, and context. Your study design should be based on sound biostatistical principles and ethical standards, and address potential sources of bias, confounding, or error. Your methods should be appropriate for the type, level, and distribution of your data, and allow you to test your hypotheses and answer your research question. You should also consider the validity and reliability of your measurement instruments, and the feasibility and practicality of your data collection and analysis procedures.

添加您的观点

3 Follow a protocol and document your steps

The third step is to follow a protocol and document your steps throughout your study. A protocol is a detailed plan that describes your study design, methods, procedures, and ethical considerations. It helps you to organize your study, ensure consistency and transparency, and avoid deviations or mistakes. You should also document your data collection, management, analysis, and reporting steps, and keep track of any changes, issues, or decisions you make along the way. This will help you to ensure accuracy and quality, and facilitate replication and verification of your results.

添加您的观点

4 Check your assumptions and results

The fourth step is to check your assumptions and results before drawing any conclusions or making any claims. You should verify that your data meet the assumptions of your chosen methods, and perform appropriate tests or corrections if needed. You should also check your results for errors, outliers, anomalies, or inconsistencies, and explore the sources and implications of these findings. You should also assess the significance, effect size, confidence intervals, and practical relevance of your results, and compare them with existing literature and expectations.

添加您的观点

Medha Sharma

Research Associate in Genomics and Bioinformatics at Celebal Technologies | Data Scientist | Deep Learning specialization | Cofounder - Constructivist | Cofounder - Silly Opera.
举报内容
Talking about Data Science in Biotechnology, it literally eases the work. Checking anomalies/outliers with Data Science is faster than traditional computational methods. It is also great comparing two methods simultaneously.

已翻译

赞
Darko Medin

Data Scientist and a Biostatistician. Developer of ML/AI models. Researcher in the fields of Biology and Clinical Research. Helping companies with Digital products, Artificial intelligence, Machine Learning.
举报内容
Well, to check the validity of the results, the results should be falsified, meaning logically tested by contradiction via empirical experiment or other empirical test. But this is not always done, anyway you should chose methods that can be falsifiable (tested as explained) in the future. Before all this you must have enough context around the data to validate the data itself. Finally dealing with sources of bias methodologically and trough data science is probably one of the most important aspects here.

已翻译

赞

5 Report your results honestly and clearly

The final step is to report your results honestly and clearly, following the ethical and regulatory guidelines of your field and journal. You should report your results in a way that is accurate, complete, transparent, and unbiased, and avoid selective reporting, data manipulation, or misinterpretation. You should also report your results in a way that is clear, concise, and coherent, and use appropriate tables, figures, and text to convey your main findings and implications. You should also acknowledge the limitations, uncertainties, and assumptions of your study, and suggest directions for future research.

添加您的观点

Nouran Hamza

Owner |Biostatistics | Public Health | Clinical Research | Host biostat done 4u podcast| Data to patient care decision.
举报内容
Reproducibility of results? Just use R, and report with R markdown so all the codes with their corresponding results will be just there! Easy beazy lemon squeezy !

已翻译

赞

6 Here’s what else to consider

This is a space to share examples, stories, or insights that don’t fit into any of the previous sections. What else would you like to add?

添加您的观点

Darko Medin

Data Scientist and a Biostatistician. Developer of ML/AI models. Researcher in the fields of Biology and Clinical Research. Helping companies with Digital products, Artificial intelligence, Machine Learning.
举报内容
In practice one can rarely prove the study result as valid, unless it's a specific example where this can be done. But if the uncertainty around validity is objectively reported and if possible decreased in favour of beyond the reasonable doubt/ highly probable valid result, this the the direction towards optimal study.

已翻译

赞
Tamara Perellón

Facultativo en CENTRE D'ANALISIS GIRONA SA
举报内容
At the biomedical lab external quality programmes are used to ensure the reproducibility and low bias on your analysis reports. Same samples are tested by an expert lab and then sent to hundreds of labs enrolled to the programme. Useful to monitorise sample treatment, the analysers performance but also the scientists training on result interpretation. Also, when you know the results reported by majority of labs, they can be used to train new colleagues or solving deviations.

已翻译

赞
Waqas Arshad Abbasi PSPO?, PSM?

The Product Guy ??
举报内容
To ensure study results are valid and reliable, focus on designing robust experiments with clear, measurable objectives, employ statistically significant sample sizes, and utilize repeatable and consistent methods across trials. Rigorous peer review and validation against established benchmarks are critical. Lastly, transparent documentation and open collaboration with the research community enhance credibility and reproducibility.

已翻译

赞

Biotechnology

+ 关注

给文章评分

我们借助人工智能创建了此文章。您认为这篇文章怎么样？

很棒不太好

举报此文章

查看全部

How can you ensure study results are valid and reliable?

1

2

3

4

5

6

1 Define your research question

2 Choose your design and methods

3 Follow a protocol and document your steps

4 Check your assumptions and results

5 Report your results honestly and clearly

6 Here’s what else to consider

Biotechnology

给文章评分

感谢您的反馈

更多Biotechnology相关文章

更多相关阅读内容