Scientific Writing Best Practices

𝗪𝗲 𝘁𝗿𝘂𝘀𝘁 𝗼𝘂𝗿 𝗠𝗜𝗦 𝗿𝗲𝗽𝗼𝗿𝘁𝘀. After all, they’re “𝘥𝘢𝘵𝘢-𝘥𝘳𝘪𝘷𝘦𝘯,” right? But here’s the uncomfortable truth: 𝗘𝘅𝗰𝗲𝗹 𝘀𝗵𝗲𝗲𝘁𝘀 𝗰𝗮𝗻 𝗯𝗲 𝗯𝗶𝗮𝘀𝗲𝗱... 𝘀𝗼𝗺𝗲𝘁𝗶𝗺𝗲𝘀 𝘀𝘂𝗯𝘁𝗹𝘆, 𝘀𝗼𝗺𝗲𝘁𝗶𝗺𝗲𝘀 𝗱𝗮𝗻𝗴𝗲𝗿𝗼𝘂𝘀𝗹𝘆. ➡️ 𝘚𝘦𝘭𝘦𝘤𝘵𝘪𝘷𝘦 𝘒𝘗𝘐s: Highlighting only the “feel-good” metrics while ignoring the ones that reveal cracks. ➡️ 𝘊𝘩𝘦𝘳𝘳𝘺-𝘱𝘪𝘤𝘬𝘦𝘥 𝘵𝘪𝘮𝘦𝘧𝘳𝘢𝘮𝘦𝘴: A report may look stellar if you show Q2, but not so much if you include Q1. ➡️ 𝘏𝘪𝘥𝘥𝘦𝘯 𝘧𝘪𝘭𝘵𝘦𝘳𝘴 𝘪𝘯 𝘗𝘪𝘷𝘰𝘵𝘛𝘢𝘣𝘭𝘦𝘴: A simple unchecked box can completely skew the story your data is telling. These aren’t just harmless quirks. They can 𝗱𝗶𝘀𝘁𝗼𝗿𝘁 𝗯𝘂𝘀𝗶𝗻𝗲𝘀𝘀 𝗿𝗲𝗮𝗹𝗶𝘁𝘆, leading to wrong calls. Whether it’s approving budgets, launching products, or restructuring teams. So, how do we keep reports honest? ✅ Cross-check KPIs against original objectives. ✅ Review multiple time horizons, not just the “best-looking” ones. ✅ Audit filters and slicers before presenting. ✅ Encourage a culture where bad news is not buried, but acted upon. Because at the end of the day, data doesn’t lie... but reporting can. The real question is: 𝘈𝘳𝘦 𝘸𝘦 𝘳𝘦𝘢𝘥𝘺 𝘵𝘰 𝘤𝘰𝘯𝘧𝘳𝘰𝘯𝘵 𝘵𝘩𝘦 𝘸𝘩𝘰𝘭𝘦 𝘵𝘳𝘶𝘵𝘩? 👉 𝙃𝙤𝙬 𝙙𝙤 𝙮𝙤𝙪 𝙚𝙣𝙨𝙪𝙧𝙚 𝙣𝙚𝙪𝙩𝙧𝙖𝙡𝙞𝙩𝙮 𝙖𝙣𝙙 𝙩𝙧𝙖𝙣𝙨𝙥𝙖𝙧𝙚𝙣𝙘𝙮 𝙞𝙣 𝙩𝙝𝙚 𝙧𝙚𝙥𝙤𝙧𝙩𝙨 𝙮𝙤𝙪 𝙬𝙤𝙧𝙠 𝙬𝙞𝙩𝙝? #DataDrivenDecisionMaking #DataAnalytics #ExcelReports #DataTransparency #MISReporting

🧠 CIA Daily Drill #020 The “Relationship Pressure” Trap 📌 CIA Syllabus: Part 1 – Domain 2: Ethics & Professionalism Part 2 – Domain 4: Communicating Results & Monitoring Progress Part 3 – Governance & Organizational Relationships 🎯 QUESTION OF THE DAY During a final audit discussion, management strongly disagreed with several significant audit observations related to weak procurement controls and inadequate monitoring activities. The audit manager believed the observations were properly supported by sufficient audit evidence. However, the audit manager also knew that the auditee executive had a close relationship with senior leadership and could negatively influence future support for the internal audit function. To avoid conflict, a senior member of the audit team suggested softening or removing some observations from the final report. What is the BEST action for internal audit management to take? A. Modify or remove the observations to preserve a positive working relationship with senior management B. Delay issuing the report until management becomes more cooperative and accepting of the observations C. Maintain the observations if they are properly supported, while communicating them objectively and professionally D. Escalate the disagreement immediately to the board because management disagreement automatically impairs audit independence 🧠 KEY INSIGHT ⚠️ This is a “relationship pressure vs audit integrity” judgment trap - Internal audit effectiveness depends on credibility, objectivity, and evidence-based reporting - Strong relationships with management should not influence audit conclusions or suppress supported observations - Professional disagreement does not automatically require escalation if observations are adequately supported and communicated appropriately - Auditors must balance professionalism and diplomacy without compromising independence or integrity 👉 Internal audit protects its long-term credibility by reporting supported issues objectively, even under political pressure. 📌 CORRECT ANSWER: C ✔ Preserves audit integrity and objectivity ✔ Relies on sufficient and appropriate evidence ✔ Maintains professional communication standards ✔ Avoids inappropriate influence from organizational relationships ❌ WHY OTHERS FAIL A → Relationship concerns should not override supported audit conclusions B → Delaying reporting does not resolve objectivity or governance concerns D → Disagreement alone does not automatically require escalation to the board 🧩 MEMORY ANCHOR 👉 Good relationships should improve communication, not weaken audit conclusions. 📢 Follow for advanced CIA drills focused on audit judgment, governance pressure, organizational politics, and real-world internal audit dilemmas. #CIAExam #InternalAudit #AuditJudgment #Governance #ProfessionalSkepticism #AuditReporting #CIAExamPrep

I tested over 100 research tools with my PhD students. Only 20 made the cut. I started keeping a list around year 8 of supervising. A student would tell me something had genuinely helped them. I would test it myself. Either it stayed on the list or it did not. 18 years later, the list is still 20. Here they are. 1) Literature Search Semantic Scholar, Lens.org, Connected Papers, Scopus free access Semantic Scholar is the one I push hardest. You can trace citations forward and backward in ways Google Scholar simply cannot. If Google Scholar is your only search tool right now, this is the first thing to change. 2) Reference Management Zotero, Mendeley, Paperpile If you are not using Zotero already, download it today. Every student I supervise uses it from month one. None have ever asked to switch away from it. Paperpile has no free version but it has a 30-day trial and a 50% academic discount after that, which brings it to around $5 a month. Worth it if you write a lot. 3) Writing and Grammar Hemingway Editor, Inkwell.net, WordCounter Hemingway shows you when your writing is more complex than it needs to be. Useful for chapters that still feel heavy after you have already edited them twice. Inkwell.net is newer but genuinely good for academic tone and clarity checks. 4) Data Analysis R, ATLAS.ti free trial, MAXQDA free trial, 5) AI Research Assistants Consensus, Elicit, Research Rabbit These three are the main AI tools I recommend for literature work right now. They cite real sources. They do not make references up. That matters more than any other feature on the list. 6) Productivity Notion, Toggl Track, Focusmate, Cold Turkey Focusmate pairs you with a real person for a 50-minute working session over video. No chat, no distraction. Just two people getting their work done at the same time. Several of my students swear by it for thesis writing days. My personal top 3? Semantic Scholar for searching. Zotero for managing references. Consensus for quick scoping when a student is new to a field and needs to map the territory fast. Start there. Add the rest when you need them. One more thing. For most students I work with, the tools are not actually the gap. The gap is knowing how to use them together in the right order. That is what most PhD programmes never teach. Which tools would you add to this list? Save this toolkit. Repost for your PhD network. #PhDTips --- If your toolkit gaps are not tools but structure and strategy, my 1:1 mentoring programme is built for working professionals doing PhDs and DBAs. Fill in the discovery form at www.phdtoprof.com

Writing a Thesis or Paper with AI Tools INTRODUCTION Purpose: Set the stage for your research Key elements Context: Provide a simple background so even non-specialists can understand Problem/Gap: Point out what is missing in existing research Purpose/Aim: State what your research will achieve Research Questions/Objectives: Exact questions you’re answering. Contribution/Significance: Show who benefits and how 👉 AI Tools (Chapter 1) Grammarly, Paperpal, Trinka.ai → Help refine writing, improve clarity, grammar,and academic tone. LITERATURE REVIEW Purpose: Show what’s already known and where gaps exist. Key elements: Themes/strands in literature: Organize by themes, not paper-by-paper. Where findings agree/disagree: Highlight consensus and debate. What remains underexplored: Identify your research gap 👉 AI Tools (Chapter 2): Elicit: AI for summarizing and extracting insights from research papers. Connected Papers: Visualize connections between studies Elsevier, Mendeley: Databases and reference managers METHODS Purpose: Explain how the study was conducted so others can replicate Key elements: Design/Approach: Qualitative, quantitative, or mixed Participants/Sampling: Who took part, how many, and how chosen Measures/Instruments: What tools or scales were used Procedure/Data Collection: How and when data was collected Analysis: Statistical or qualitative analysis. Ethics: Approvals and consent. 👉 AI Tools (Chapter 3): draw.io: Visual diagrams of research design. Grok: AI reasoning and data structure support. RESULTS Purpose: Present findings clearly. Key elements: Primary outcomes: Link to each research question. Key statistics: Effect sizes, confidence intervals, p-values. Tables/Figures: To simplify complex results. 👉 AI Tools (Chapter 4): Python, MATLAB Coding: For statistical analysis. BioRender, Figma: Create visuals and scientific figures. MAXQDA, NVivo: Qualitative data analysis software. SPSS: Widely used for quantitative analysis. DISCUSSION Purpose: Interpret results and explain meaning. Key elements: Answer research questions. Compare with prior studies. Explain mechanisms (why results happened). Implications: Impact on theory, practice, or policy. Limitations: Be transparent about weaknesses. Future research: Suggest next steps. 👉 AI Tools (Chapter 5): Writefull, Trinka.ai: Improve academic language. SPSS: Further statistical interpretations. CONCLUSION Purpose: Wrap up the study in a concise way. Key elements: Recap: Purpose, method, and findings in one paragraph. Contribution: Show how gap was addressed. Limitations + Future Work: Mention briefly. Closing sentence: Strong final statement. 👉 AI Tools (Chapter 6): Grok, Perplexity, SciSpace: Generate summaries, simplify complex findings, and support critical discussion. REFERENCES Purpose: Credit sources properly. Must follow citation styles (APA, MLA, Chicago, etc.). 👉 AI Tools: Zotero, RefWorks, Mendeley, Elsevier: For managing references and automatic citation formatting.

Most medical writers are still using ChatGPT for everything. Meanwhile, they're missing specialized AI tools that actually understand scientific literature. Here are 3 tools that changed my literature review process completely: 📌 Consensus AI This one searches through 200+ million research papers and gives you instant answers backed by actual studies. It even shows you if there's scientific consensus on your question or if researchers are still debating it. 📌 Elicit Think of it as your research assistant that reads papers for you. It extracts key findings, methodology, and results into clean tables. I used to spend hours reviewing papers. Now? It happens in minutes. 📌 Scite AI This tool shows you how papers have been cited – and whether those citations support or contradict the findings. Amazing for checking if that 2019 study you're referencing has been debunked or validated since. ChatGPT is good for drafting and brainstorming. But these tools are built specifically for scientific literature. They understand study design, citation context, and research methodology in ways general AI doesn't. I still fact-check everything (always will). Always have to! But my literature reviews take significantly less time now. P.S. I also used Open Evidence before. However, it's no longer available in India. You can still use VPN and use it. But I just find these 3 more user friendly now. Have you used any of these before? #medicalwriting #medicalcopywriting #medcomms #healthcomms #medicalcommunication #healthcommunication #scientificwriting #freelance #freelancemedicalwriting #medicomarketing #mbbs #neetpg #usmle #nonclinical #pharma #biotech #medicaldevices

PhDs feeling overwhelmed while writing your thesis or manuscript? Writing is about clarity, structure, precision and time. And for many PhDs, that last bit is always in short supply. That’s where AI tools come in. But here's the truth no one tells you: AI won't write your thesis for you. But it can make your writing process 5x faster and 10x smoother, if you use the right tool, the right way. After mentoring researchers and testing tools myself, here is how I recommend you use AI effectively and ethically for your academic writing journey: 1. Don’t start with a blank page. Use tools like ChatGPT or Jenni AI to break the ice. Ask for a draft structure, a list of ideas or even a sample paragraph to get started. Think of it as a brainstorming buddy, not a ghostwriter. Tip: Feed it your research objective or abstract and ask, “Can you suggest a logical flow for this section?” 2. Let AI sharpen your tone, not dilute your voice. Tools like Wordtune, QuillBot and Trinka.ai are excellent for paraphrasing and tone adjustments. But always cross-check the rewritten content. Your voice matters. Don’t let it sound robotic. 3. Know when to switch gears from ChatGPT to academic-specific tools. Generic AI tools are great, but when you’re aiming for journal-level quality, Paperpal, WriteWise or SciSpace AI Writer offer suggestions tailored to academic writing, like fixing passive voice in methods sections or checking for field-specific terminology. 4. Edit with the eyes of a reviewer. Use ProWritingAid or Hemingway Editor for readability checks. These won’t help with your literature review content, but they will ensure your long-winded paragraph isn’t scaring readers away. Tip: Look at the readability score. Aim for clarity, not complexity. 5. Avoid AI overkill. You don’t need all 15 tools. Pick 2–3 that suit your workflow: Brainstorming: ChatGPT or Writesonic Paraphrasing & Tone: Quillbot, Wordtune, Jenni Academic-specific feedback: Paperpal, SciSpace AI, WriteWise Grammar/Plagiarism: Trinka ✅ Start with what’s free. Upgrade only if you feel a tool saves you real time or improves actual quality. One last expert insight: AI is only as smart as the prompts you give it. So ask better questions. And never skip the final human edit. PS: Which AI tools have you tried for academic writing? Any that saved your time (or gave you a headache)? Share in the comments! REPOST to help others.

Lesson Learned: "Don't Drink Client's Bathwater" As a consultant, have you ever found yourself so immersed in developing solid client relationship that you started seeing the world through their eyes? During my consulting days, I had an experience that reminded me of the importance of maintaining an objective perspective, no matter how compelling a client's narrative might be. I was working with a mining client (my first project) where we were trying to optimise cost of operations. My client was passionate, driven, and had deep expertise in the domain, naturally, it created an amazing impression in my mind and I found myself wholeheartedly buying into their narrative / perspective, advocating for their judgement without questioning it. But then, reality hit. 🚨 We started noticing discrepancies between their performance and the market benchmarks. We were nowhere near the benchmarks. It became clear that in my eagerness to support the client, I had overlooked critical signs and failed to challenge their assumptions. Here are three key takeaways from this experience: 1️⃣ Maintain Objectivity: While it's important to support your client, it's equally important to provide honest, objective feedback based on data and market insights. 2️⃣ Challenge Assumptions: Don’t be afraid to question and challenge your client constructively. This isn’t about being negative but about ensuring that the strategy is grounded in reality. 3️⃣ Stay True to Data: Let data be your guide. In the end, facts and figures should drive decisions. Ensure that your ideas are data-backed and not solely based on someone's "experience". This experience was a valuable reminder that while it's important to be passionate and supportive, we must also remain critical thinkers and advisors. "Don't drink client bathwater" means keeping your objectivity intact and ensuring that you're always providing the best possible guidance to your clients. Has anyone else experienced something similar? How do you balance supporting your client with maintaining objectivity? Let's discuss! 💬 #ClientManagement #ConsultingLife #ProfessionalGrowth

My Master’s Research Toolkit: Tools That Can Save HOURS in Research During my master’s research journey, I used a set of tools that truly made my life easier. If you are in the middle of your research or about to start, these tools can help you stay organised, and even spark creativity. Here’s what I used and why I loved them: 1️⃣ 𝐑𝐞𝐟𝐞𝐫𝐞𝐧𝐜𝐞 𝐌𝐚𝐧𝐚𝐠𝐞𝐦𝐞𝐧𝐭 ➤ 𝐙𝐨𝐭𝐞𝐫𝐨: My go-to tool for managing and citing references. It organises research papers seamlessly and generates citations in any format. Link - https://www.zotero.org/ ➤ 𝐆𝐨𝐨𝐠𝐥𝐞 𝐒𝐜𝐡𝐨𝐥𝐚𝐫: Essential for finding scholarly articles, tracking citations, and discovering the latest research. Link - https://lnkd.in/ez_pbD9J ➤ 𝐒𝐞𝐦𝐚𝐧𝐭𝐢𝐜 𝐒𝐜𝐡𝐨𝐥𝐚𝐫: A smarter way to search and analyse academic papers with AI-powered insights. Link - https://lnkd.in/eQ4CXpEj 2️⃣ 𝐖𝐫𝐢𝐭𝐢𝐧𝐠 & 𝐅𝐨𝐫𝐦𝐚𝐭𝐭𝐢𝐧𝐠 ➤ 𝐎𝐯𝐞𝐫𝐥𝐞𝐚𝐟: Perfect for collaborative writing in LaTeX, especially when working on complex research papers. Link - https://lnkd.in/eupumZXu ➤ 𝐆𝐫𝐚𝐦𝐦𝐚𝐫𝐥𝐲: Ensures that my writing is professional, error-free, and concise. Link - https://app.grammarly.com/ 3️⃣ 𝐑𝐞𝐬𝐞𝐚𝐫𝐜𝐡 𝐃𝐢𝐬𝐜𝐨𝐯𝐞𝐫𝐲 & 𝐌𝐚𝐧𝐚𝐠𝐞𝐦𝐞𝐧𝐭 ➤ 𝐑𝐞𝐬𝐞𝐚𝐫𝐜𝐡 𝐑𝐚𝐛𝐛𝐢𝐭: A powerful tool for discovering related research papers and expanding your knowledge network. Link - https://lnkd.in/e52ZEnfX ➤ 𝐍𝐨𝐭𝐢𝐨𝐧: My productivity hub for organising tasks, notes, and project timelines. Link - https://www.notion.so/ ➤ 𝐎𝐛𝐬𝐢𝐝𝐢𝐚𝐧: Perfect for building a second brain with linked notes and structured knowledge. Link - https://obsidian.md/ 4️⃣ 𝐕𝐢𝐬𝐮𝐚𝐥𝐢𝐬𝐚𝐭𝐢𝐨𝐧 & 𝐃𝐞𝐬𝐢𝐠𝐧 ➤ 𝐂𝐚𝐧𝐯𝐚: For creating visually appealing presentation slides and posters. Link - https://www.canva.com/ ➤ 𝐃𝐫𝐚𝐰.𝐢𝐨: My favourite for creating flowcharts and diagrams to visualise processes and concepts. 5️⃣ 𝐏𝐫𝐨𝐠𝐫𝐚𝐦𝐦𝐢𝐧𝐠 & 𝐃𝐞𝐯𝐞𝐥𝐨𝐩𝐦𝐞𝐧𝐭 ➤ 𝐕𝐢𝐬𝐮𝐚𝐥 𝐒𝐭𝐮𝐝𝐢𝐨: A versatile IDE for coding and debugging projects. Link - https://lnkd.in/ekyBuP9d ➤ 𝐆𝐢𝐭𝐇𝐮𝐛: Essential for version control and collaboration on research code. Link - https://github.com/ 6️⃣ 𝐏𝐫𝐨𝐝𝐮𝐜𝐭𝐢𝐯𝐢𝐭𝐲 & 𝐅𝐨𝐜𝐮𝐬 ➤ 𝐏𝐨𝐦𝐨𝐝𝐨𝐫𝐨 𝐓𝐢𝐦𝐞𝐫𝐬: Helped me stay productive with focused work intervals and regular breaks. Link - https://pomofocus.io/ 7️⃣ 𝐀𝐝𝐯𝐚𝐧𝐜𝐞𝐝 𝐓𝐨𝐨𝐥𝐬 𝐟𝐨𝐫 𝐑𝐞𝐬𝐞𝐚𝐫𝐜𝐡𝐞𝐫𝐬 ➤ 𝐒𝐜𝐢𝐒𝐩𝐚𝐜𝐞: Simplifies formatting research papers to meet journal requirements with minimal effort. Link - https://typeset.io/ Have you tried any of these? Or do you know a tool that helped in your research? SHARE in the comments! If this list helped you or could help someone you know, please LIKE, SHARE, or REPOST it. Let’s make research a little less overwhelming for everyone. #ToolsForResearchers #AcademicWriting #OrganizedResearch

📜 🖇 Report writing is the window into your competence as an investigator and a potential goldmine of opportunities for those looking to raise doubt about your competence or even your integrity as an investigator. None of us wants to have our work characterized as poor, biased, unfair, or incompetent simply because we didn't manage our writing and ensure it stood up to review. It isn't about winning a Giller or Pulitzer prize, but it needs to be solid. Here are a few red flags to avoid in your reports: 🚩 Unfocused Objectives: A well-defined mandate is critical to understanding the scope and nature of the investigation. This needs to be clearly stated at the outset of your report. 🚩Vague or Missing Methodology: An investigation is as much about your process as the outcome. Ensure you provide the investigation steps, including what you did and did not do. This helps readers understand your approach and will instill confidence, lending credibility to your process. 🚩 Loose Timelines: Understanding the timeline of complex issues or events is often crucial to readers grasping the nuances of what was investigated. Make this clear, and consider a visual aid for the timeline for complex matters. 🚩 Scant Evidence: In your efforts to support your findings, or in the case of professional regulatory investigations, instilling confidence in those charged with making decisions, ensure you have a proper overview of the evidence. Remember, evidence leads to conclusions, not the other way around. Be sure anyone reading your report will see the evidence and be able to draw the logical line from that evidence to a defensible conclusion. 🚩 Context Neglected: Don't lose sight of the context around a matter. Issues described in a vacuum can be misunderstood in terms of their significance. Consider legislation, broader policies, and industry standards where appropriate. Provide the backstory to help readers understand the bigger picture. 🚩 Unintended Signals of Bias: Bias can sneak up on you. Be vigilant about your own bias. It exists in all of us. Ensure your writing is impartial. Avoid the adjectives that add emphasis to a statement. It is better to say, "The evidence supports the Complainant's assertion that..." Rather than say, "The overwhelming body of evidence clearly and definitively supported the Complainant's assertion that..." 🚩 Jargon Overload: While we live in a technical world, remember to make your reports accessible to all readers by using clear and straightforward language. 🚩 Inadequate Review: Always double and triple-check your work, and if possible, have it reviewed by someone, not just for grammar and spelling but to ensure completeness, readability, and understanding. There is more, but not in 3000 characters! Let's get another dialogue started so we all learn from each other. #Investigation #ReportWriting #BestPractices

An established campaign should not be prioritized over a new test that is clearly delivering better results. Recently, we reviewed a Google/PMax account where newer asset groups within existing campaigns began outperforming the legacy structure. The performance difference was significant. The legacy asset group had a longer track record and felt safer due to its familiarity within the account. However, the newer assets demonstrated a much clearer positive trend. This does not mean making abrupt changes after only a day or two. New campaigns and asset groups require sufficient time, budget, and conversion data to demonstrate their effectiveness. At the same time, do not disregard positive results from new tests simply because the existing campaign has more history. This is why reporting windows are important. A 14-day reporting window may underrepresent new tests if the assets were not active for the entire period. A 7-day window can provide a clearer view of current momentum. Neither window offers a complete picture, so context is essential for accurate interpretation. Teams often become attached to campaigns that were previously successful. And I get it. History feels safe. However, factors such as product mix, creative assets, search behavior, promotions, and margins all change over time. What was effective six months ago may no longer be the best use of budget today. The practical approach is to avoid ending new tests prematurely and to refrain from supporting outdated structures out of habit. Evaluate the data objectively, review the appropriate reporting window, and ensure the test has sufficient budget. When the results are strong enough, allocate budget to what is currently performing best. Forget about preserving legacy campaigns, the real win is ensuring the account continues to progress in the right direction.