General Productivity¶

Prompt:

I need your help planning my day tomorrow.
First, please list all my scheduled meetings and deadlines from my calendar.
Then, taking into account 2 hours of break time, could you suggest a time-managed itinerary for me?
Lastly, based on the priority of the tasks, can you create a task list in the order I should tackle them?

Response:

I would be happy to help you plan your day, but I need to clarify something first: I do not currently have access to your calendar or scheduling system.

To help you effectively, please share: - Your scheduled meetings and their times - Any deadlines you have tomorrow - Tasks you need to complete - Your preferred working hours (e.g., 8 AM - 6 PM)

Once you provide this information, I will:

1. Organize your schedule with meetings as fixed anchor points
2. Create a time-blocked itinerary that includes your 2 hours of break time strategically placed (ideally after cognitively demanding tasks and around lunch)
3. Prioritize your tasks using the Eisenhower Matrix approach (urgent/important, important/not urgent, etc.)

Alternatively, if you are using a platform with calendar integration enabled (such as Microsoft Copilot with Outlook, Google Gemini with Google Calendar, or a ChatGPT calendar plugin), you can use this same prompt there for automatic calendar access.

Would you like to share your schedule details so I can create a personalized plan?

Scenario: You are designing a mixed-methods study on faculty adoption of AI tools in teaching and want Claude to help you think through potential methodological challenges.

Prompt:

I'm designing a mixed-methods study to investigate how faculty at research universities
are integrating generative AI tools into their teaching practices. The study will combine
surveys with semi-structured interviews.

Please think carefully about the methodological challenges I might face and how to
address them. Consider issues of sampling, bias, validity, and the rapidly changing
nature of AI technology.

Claude's Extended Thinking Process (visible to you):

Let me think through this systematically...

First, sampling challenges: Faculty who respond to surveys about AI might be those already interested in technology - this creates self-selection bias. Also, "research universities" varies widely - R1 vs R2, public vs private, STEM-heavy vs liberal arts...

The temporal dimension is tricky. AI tools are evolving so rapidly that practices from even 6 months ago may be obsolete. Interview data collected in January might not reflect practices by the time the study is published...

Validity concerns: Faculty might overreport AI use due to social desirability, or underreport due to institutional policies they're uncertain about...

Claude's Response:

Your study design faces several interconnected methodological challenges. Here is my analysis with recommendations:

1. Sampling and Selection Bias

• Challenge: Faculty who volunteer for AI-related studies likely skew toward early adopters, creating non-representative samples
• Mitigation: Use stratified sampling across disciplines, rank, and known technology adoption patterns. Consider incentivizing participation from technology-hesitant faculty. Partner with faculty development offices who can facilitate broader recruitment

2. Temporal Validity

• Challenge: The AI landscape evolves monthly. Data collected now may not reflect practices 6-12 months later when you publish
• Mitigation: Explicitly timestamp your data collection period. Frame findings as capturing a specific moment in a trajectory. Consider a longitudinal design with follow-up interviews. Focus on underlying adoption patterns rather than specific tool usage

3. Social Desirability and Institutional Climate

• Challenge: Faculty may underreport AI use if uncertain about institutional policies, or overreport to appear innovative
• Mitigation: Ensure anonymity in surveys. In interviews, normalize both adoption and non-adoption. Include questions about institutional climate and policy clarity. Consider asking about colleague practices (indirect questioning)

4. Defining Key Constructs

• Challenge: "Integration into teaching" could mean anything from using ChatGPT for lecture prep to teaching students prompt engineering
• Mitigation: Develop a clear taxonomy of AI use in teaching. Pilot your definitions with faculty before full deployment. Include behavioral anchors in survey questions

5. Mixed-Methods Integration

• Recommendation: Use an explanatory sequential design - let quantitative findings guide interview protocol development. This allows you to probe unexpected survey results and achieve genuine triangulation

Would you like me to elaborate on any of these areas or help develop specific survey items?

Scenario: You're preparing a literature review for your dissertation on the effectiveness of peer-led instruction in undergraduate STEM courses.

Research Question:

What evidence exists about the effectiveness of peer-led instruction (also called
peer-led team learning or PLTL) in improving learning outcomes for underrepresented
students in undergraduate STEM courses, particularly in gateway courses like
introductory biology, chemistry, and physics?

What Scholar Labs Does:

Instead of searching for papers that simply contain "peer-led instruction" and "underrepresented students," Scholar Labs:

1. Analyzes the question to identify multiple dimensions:
2. Instructional method: peer-led instruction, PLTL, peer learning
3. Population: underrepresented students, diversity in STEM
4. Context: undergraduate, gateway courses, introductory science

5. Outcomes: learning effectiveness, academic performance, retention

6. Searches comprehensively across these dimensions simultaneously

7. Surfaces relevant papers even if they use different terminology (e.g., "supplemental instruction," "peer teaching," "collaborative learning")

Example Results:

Scholar Labs might return papers with contextual descriptions like:

• "Peer-Led Team Learning in General Chemistry: Implementation and Evaluation" (Wilson & Varma-Nelson, 2016) This paper reports on PLTL effectiveness in a gateway chemistry course and includes disaggregated data showing differential impacts for first-generation college students.

• "Supplemental Instruction and the Performance of Hispanic Students in Developmental Mathematics" (Peterfreund et al., 2007) While using different terminology ("supplemental instruction"), this paper addresses peer-led learning with a specific focus on Hispanic students in a foundational STEM course.

• "The Role of Near-Peer Mentoring in STEM Persistence for Underrepresented Students" (Johnson & Stage, 2018) This study examines peer mentorship structures in biology courses, providing insights into mechanisms that make peer-led instruction effective for underrepresented students.

Follow-up Questions:

Scholar Labs allows you to ask follow-up questions to narrow or expand your search:

Which of these studies include control groups and experimental designs rather than
observational studies?

Are there any meta-analyses or systematic reviews synthesizing this evidence?

What critiques or limitations have researchers identified about peer-led instruction
for diverse student populations?

To illustrate the differences between these research tools, here's a side-by-side comparison using the same research question across four different platforms: Scholar Labs, ChatGPT Deep Research, Gemini Deep Research, and Claude.

Research Question:

Find recent and highly cited peer-reviewed articles from top-ranked journals that
report the rate of polar ice melt.

Google Scholar Labs Results¶

Processing time: Instant (seconds) Output: 10 highly cited papers with contextual summaries

Scholar Labs returned highly relevant papers with citation metrics and contextual summaries explaining why each is relevant to the query:

Key papers found:

• Otosaka et al. (2023) - Mass balance of the Greenland and Antarctic ice sheets from 1992 to 2020 Earth System Science Data · Cited by 196 Mass loss accelerated from 105 Gt/yr (1992-1996) to 372 Gt/yr (2016-2020)

• Edwards et al. (2021) - Projected land ice contributions to twenty-first-century sea level rise Nature · Cited by 441 Limiting warming to 1.5°C would halve land ice contribution from 25 to 13 cm by 2100

• Hanna et al. (2024) - Short-and long-term variability of the Antarctic and Greenland ice sheets Nature Reviews Earth & Environment · Cited by 68 Total polar ice loss: -382 ± 42 Gt/year (2002-2022) = 1.1 mm/year SLE

• Greene et al. (2024) - Ubiquitous acceleration in Greenland Ice Sheet calving from 1985 to 2022 Nature · Cited by 35 AI-powered terminus mapping reveals 20% underestimate of Greenland mass loss

• Millan et al. (2023) - Rapid disintegration and weakening of ice shelves in North Greenland Nature Communications · Cited by 35 North Greenland ice shelves lost 35% of volume since 1978

Strengths:

• ✓ Instant results (seconds)
• ✓ Citation counts immediately visible ("Cited by 441")
• ✓ Contextual summaries explain why each paper is relevant
• ✓ Direct links to PDFs and institutional library access
• ✓ 100% peer-reviewed academic sources
• ✓ Clean interface for literature discovery

Gemini Deep Research Results¶

Processing time: ~5-15 minutes

Output: Comprehensive research report (~3,500 words)

View Full Gemini Deep Research Report

Report Structure:

• Executive Abstract: High-level synthesis of cryospheric state and committed changes
• Methodological Convergence: Detailed explanation of GRACE, altimetry, and input-output methods
• Greenland Analysis: Greene et al. 20% underestimate, 2023 vs 2024 atmospheric variability, firn degradation
• Antarctic Analysis: WAIS committed collapse (van den Akker et al.), Thwaites tipping point, sea ice regime shift
• Global Teleconnections: AMOC impacts, New Zealand climate effects, Southern Hemisphere circulation
• Comparative Tables: Greenland metrics (2023-2025), polar ice sheet dynamics comparison
• 25+ Citations: Nature, Science, The Cryosphere, PNAS, plus news and policy sources

Strengths:

• ✓ Formal research report structure with executive abstract
• ✓ 25+ sources including academic papers, government reports, and news
• ✓ Mathematical equations and technical depth
• ✓ Comparative tables and data synthesis
• ✓ Broader context (AMOC impacts, teleconnections, New Zealand climate)
• ✓ Methodological explanations (GRACE vs. altimetry)
• ✓ Export-ready format for grant proposals

Claude Results¶

Processing time: 5-10 minutes (this example took 6 minutes)

Output: Annotated bibliography with 23 papers organized thematically

View Full Claude Annotated Bibliography

Content Organization:

• Greenland Ice Sheet Studies (7 papers): IMBIE 2020/2023, Greene et al. calving acceleration, Box et al. committed losses, Briner et al. Holocene context
• Antarctic Ice Sheet Studies (6 papers): Smith et al. 2020, Milillo et al. rapid retreat, Naughten et al. unavoidable warming, Schmidt et al. Thwaites robotic observations
• Sea Ice Studies (7 papers): Arctic ice-free projections (Jahn et al.), Antarctic regime shift (Purich & Doddridge), 2023 record lows (Josey et al.)
• Sea Level Rise Studies (3 papers): Edwards et al. projections, Bamber et al. expert elicitation, IMBIE 2018

Strengths:

• ✓ Most papers found (23 comprehensive sources)
• ✓ Excellent thematic organization (Greenland → Antarctica → Sea Ice → SLR)
• ✓ Citation counts + DOIs + impact metrics included
• ✓ Balanced technical detail with accessibility
• ✓ 100% peer-reviewed academic sources
• ✓ Fast comprehensive synthesis

ChatGPT Deep Research Results¶

Processing time: ~3 minutes

Output: Structured synthesis (~1,000 words) with in-text citations

View ChatGPT Deep Research Results

Key findings presented:

• Greenland: 3,900 ± 340 Gt lost (1992-2018), peaking at 345 ± 66 Gt/yr in 2011
• Antarctica: Loss grew six-fold from ~40 Gt/yr (1979-1990) to 252 Gt/yr (2009-2017)
• Combined: Ice loss quadrupled since 1990s, rising from 105 Gt/yr to 372 Gt/yr (2016-2020)
• Methods: IMBIE consortium data fusion of altimetry, gravimetry, and ice-velocity mapping

Strengths:

• ✓ Clear numerical data presentation with uncertainties
• ✓ Strong focus on consensus findings (IMBIE consortium)
• ✓ In-text citations with superscripts
• ✓ Accessible academic writing style
• ✓ Methods explained for each study
• ✓ 10 peer-reviewed sources from Nature, Science, PNAS
• ✓ Concise synthesis format

Four-Way Comparison¶

Papers all four tools found:

• ✓ Otosaka et al. (2023) - IMBIE mass balance study (1992-2020)
• ✓ Greene et al. (2024) - Greenland calving acceleration (20% underestimate)
• ✓ Edwards et al. (2021) - Land ice projections
• ✓ Smith et al. (2020) - Science, ICESat altimetry

Unique contributions by tool:

• Scholar Labs: Best citation metrics ("Cited by 441"), fastest access to papers, institutional library links
• ChatGPT: Strong IMBIE consortium focus, clear numerical consensus, concise synthesis (~1,000 words)
• Gemini: Most comprehensive coverage, news sources (Guardian), policy reports (NOAA), teleconnection studies, formal report structure (~3,500 words)
• Claude: Most papers (23), sea ice regime shift studies, expert elicitation (Bamber et al.), thematic organization

Workflow Recommendation¶

For a comprehensive research project, use multiple tools strategically:

Quick Start (10 minutes total):

1. Scholar Labs (instant) → Get 10 high-impact papers with citation counts
2. Claude (30 seconds) → Get 23 papers organized thematically
3. ChatGPT Deep Research (3 min) → Get consensus numerical data and IMBIE focus

Comprehensive Research (15-20 minutes total):

1. Scholar Labs (instant) → Identify highest-impact papers with institutional access
2. Claude (30 seconds) → Get broad academic coverage across all aspects
3. ChatGPT Deep Research (3 min) → Get clear consensus findings with uncertainties
4. Gemini Deep Research (5 min) → Get formal report with news, policy, and broader context

Result: 35+ unique sources spanning peer-reviewed literature, policy reports, news coverage, and consensus data—ready for synthesis into grant proposals, literature reviews, or background sections.

Best tool for specific needs:

• Need papers NOW? → Scholar Labs
• Writing a grant background? → Gemini Deep Research
• Need consensus numbers? → ChatGPT Deep Research
• Comprehensive literature review? → Claude

Use Extended Thinking when:

• You have a complex analytical question: "What are the trade-offs between different approaches to measuring student engagement in online courses?"
• You need to work through methodology: "Help me think through the validity threats to my proposed quasi-experimental design."
• You want transparent reasoning: "Analyze the arguments for and against requiring AI literacy in general education."

Use Deep Research when:

• You need current information: "What are the latest developments in AI-assisted grading tools?"
• You want comprehensive coverage: "Survey the landscape of open educational resources for teaching data science."
• You need citations and sources: "Find recent studies on the effectiveness of flipped classroom approaches in STEM education."

Use Scholar Labs when:

• You need peer-reviewed academic literature: "Find recent research on the rate of polar ice melt from top-ranked journals."
• Your research question has multiple dimensions: "What evidence exists about peer-led instruction effectiveness for underrepresented students in STEM gateway courses?"
• You want papers that address your specific question rather than just containing keywords: "How do community colleges support adult learners returning to higher education after career changes?"
• You need to explore connections across subfields: "What interdisciplinary research exists on climate change communication in K-12 education?"

Combine all three when:

• Start with Scholar Labs to find relevant peer-reviewed literature on your topic
• Use Deep Research to gather broader context, including grey literature, policy documents, and recent developments not yet in academic journals
• Apply Extended Thinking to synthesize findings, identify gaps, and develop your research questions or methodology
• Example workflow: Use Scholar Labs to find studies on AI in peer review → Use Deep Research to survey publisher policies and recent news → Use Extended Thinking to analyze the ethical implications and design interview questions for journal editors