Manuel B. Garcia

Manuel B. Garcia serves as the Senior Director for Educational Technology and Digital Learning at FEU Institute of Technology, Manila, Philippines. Read More

Contact Info

1607, FEU Tech Building,
P. Paredes St, Sampaloc,
Manila, Philippines
mbgarcia@feutech.edu.ph

Follow Me

The Complete Library

All Research Guides

Browse every published guide in the Researcher's Field Guide, including the complete research journey and supplemental collections, listed alphabetically for quick reference.

Complete Library

Every published guide

Showing 701–750 of 5,065 guides

5,065 guides
PART 0D · SUBPART 33 · GUIDE 495

Verifying AI-Generated Research Figures

How Do Researchers Verify Every Value in an AI-Generated Research Figure?

An AI-generated figure may look accurate while displaying incorrect values, labels, or statistical summaries. Learn a systematic verification process that traces every plotted element back to the underlying research data.

PART 0D · SUBPART 33 · GUIDE 494

Misleading Visual Emphasis in AI Figures

Can Generative AI Unintentionally Create Misleading Visual Emphasis?

An AI-generated research figure can display correct data yet communicate a misleading conclusion. Learn how visual scale, color, positioning, and selective emphasis can distort the interpretation of scientific evidence.

PART 0D · SUBPART 33 · GUIDE 493

Disclosing AI-Generated Graphical Abstracts

Should AI-Generated Graphical Abstracts Be Disclosed?

AI assistance in graphical abstracts may require disclosure, but transparency alone does not establish permission to publish AI-generated artwork. Learn how publisher policies differ and what researchers should document.

PART 0D · SUBPART 33 · GUIDE 492

AI-Generated Graphical Abstracts

Can Generative AI Create a Graphical Abstract for Research?

Generative AI can help researchers turn a study into a graphical abstract, but visual appeal does not guarantee scientific accuracy. Learn how to create, verify, and prepare AI-assisted graphical abstracts for publication.

PART 0D · SUBPART 33 · GUIDE 491

AI-Generated Conceptual Research Diagrams

Can Generative AI Create a Conceptual Research Diagram?

Generative AI can help researchers translate concepts, variables, and proposed relationships into visual diagrams. However, an attractive conceptual model is not necessarily a theoretically defensible one.

PART 0D · SUBPART 33 · GUIDE 490

AI for Scientific Figure Labels and Captions

Can Generative AI Help Improve Scientific Figure Labels and Captions?

Generative AI can improve the clarity, consistency, and readability of scientific figure labels and captions. However, seemingly minor wording changes can misrepresent statistical results, measurement scales, or causal relationships. Learn how to use AI for figure editing while preserving the precise meaning of the research evidence.

PART 0D · SUBPART 33 · GUIDE 489

AI-Invented Data in Research Visualizations

Can Generative AI Invent Data While Creating a Research Visualization?

Generative AI can introduce observations, statistics, or trends that were never present in a research dataset. Learn how invented values enter visualizations, why synthetic data are not necessarily misconduct, and how researchers can prevent fabricated evidence from appearing in scientific figures.

PART 0D · SUBPART 33 · GUIDE 488

AI Graphs With Incorrect Data Values

Can Generative AI Create a Graph That Looks Correct but Plots the Wrong Values?

AI-generated graphs can appear accurate while displaying incorrect values, mislabeled categories, or improperly calculated statistics. Learn how these errors arise, why visual inspection may not reveal them, and how researchers can verify the numerical integrity of every plotted element.

PART 0D · SUBPART 33 · GUIDE 487

AI Figures From Raw Research Data

Can Generative AI Create an Accurate Figure From Raw Research Data?

Generative AI can transform raw research data into figures using plotting code and analytical tools. However, accurate figures require verified data transformations, correct statistical calculations, and faithful graphical encoding. Learn how to establish a reproducible workflow and audit every stage from source data to final figure.

PART 0D · SUBPART 33 · GUIDE 486

AI for Choosing Research Graphs

Can Generative AI Choose an Appropriate Graph for Research Data?

Generative AI can recommend graphs based on variable types, research questions, and analytical goals. However, an appropriate visualization also depends on the study design, distribution, statistical analysis, and intended inference. Learn how to evaluate AI chart recommendations and avoid graphs that conceal or distort research findings.

PART 0D · SUBPART 33 · GUIDE 485

AI-Generated Research Figures and Graphs

Can Generative AI Create Research Figures and Graphs?

Generative AI can help researchers create graphs, diagrams, and scientific figures, but visual plausibility does not guarantee accuracy. Learn how to distinguish code-generated visualizations from AI-rendered images, verify plotted values, and document AI assistance in research figures.

PART 0D · SUBPART 32 · GUIDE 484

AI Image Enhancement and Research Falsification

When Does AI Image Enhancement Cross Into Research Falsification?

AI image enhancement may cross into research falsification when it introduces, removes, or changes scientific evidence so that the research record no longer accurately represents the observations. Learn how to distinguish legitimate processing, methodological error, questionable practices, and formal misconduct.

PART 0D · SUBPART 32 · GUIDE 483

Disclosing AI Image Processing

How Should AI Image Processing Be Disclosed in Research?

Disclosing AI image processing requires more than naming the tool. Learn how to report the purpose, methods, transformations, validation, and limitations of AI-assisted scientific images while meeting applicable journal requirements.

PART 0D · SUBPART 32 · GUIDE 482

Preserving Original Scientific Images

Should Researchers Always Preserve the Original Scientific Images?

Original scientific images provide the evidential record needed to verify processing, measurements, and published findings. Learn what researchers should preserve, why exported images may not be sufficient, how to manage AI-generated modifications, and what determines appropriate retention periods.

PART 0D · SUBPART 32 · GUIDE 481

Validating AI Scientific Image Analysis

How Do Researchers Validate AI-Assisted Scientific Image Analysis?

Validating AI-assisted scientific image analysis requires more than checking whether processed images look accurate. Learn how to evaluate source data, segmentation performance, measurement errors, dataset shift, and the reliability of conclusions derived from AI-analyzed images.

PART 0D · SUBPART 32 · GUIDE 480

AI Segmentation as Research Measurement

Can AI-Generated Image Segmentations Be Treated as Research Measurements?

AI segmentation masks can be used to calculate research measurements, including object counts, areas, volumes, and structural dimensions. However, automated values are not automatically valid. Learn how segmentation error propagates into measurements, why overlap accuracy may be insufficient, and how to establish that AI-derived results are scientifically defensible.

PART 0D · SUBPART 32 · GUIDE 479

AI for Scientific Image Segmentation

Can AI Be Used to Segment Scientific Images?

AI can segment scientific images by identifying cells, structures, tissues, and other regions of interest. However, accurate-looking segmentation masks may contain missed objects, incorrect boundaries, or systematic errors. Learn how segmentation methods work, what researchers should validate, and when AI-generated masks can support scientific analysis.

PART 0D · SUBPART 32 · GUIDE 478

AI Denoising and Scientific Data Integrity

Can AI Denoising Alter Scientifically Meaningful Information?

AI denoising can improve the visibility of scientific images, but it may also suppress genuine weak signals, alter boundaries, or introduce misleading structures. Learn why noise removal is not necessarily neutral, how denoising affects measurements, and what researchers should validate before using processed images as evidence.

PART 0D · SUBPART 32 · GUIDE 477

AI Upscaling and Invented Image Details

Can AI Upscaling Invent Visual Details That Were Never Present in the Original Image?

AI upscaling can make scientific images appear sharper while introducing details that were not established by the original measurements. Learn how super-resolution works, why realistic-looking structures may be unreliable, and how researchers can validate upscaled images before using them as evidence.

PART 0D · SUBPART 32 · GUIDE 476

AI Filling Missing Scientific Image Details

Can AI Fill Missing Parts of a Scientific or Research Image?

AI can fill missing regions of scientific images with convincing visual details, but generated pixels are not necessarily recovered observations. Learn when model-based reconstruction may be scientifically useful, how to distinguish estimates from authentic data, and why unvalidated image completion can compromise research integrity.

PART 0D · SUBPART 32 · GUIDE 475

AI Object Removal in Scientific Images

Can AI Remove Objects or Artifacts From Scientific Images?

AI can remove visible objects and apparent artifacts from scientific images, but doing so may erase genuine evidence or replace observed information with generated pixels. Learn how to distinguish validated artifact correction from misleading object removal and determine when an edited image is unsuitable for research reporting.

PART 0D · SUBPART 32 · GUIDE 474

Scientific Image Processing vs. AI Manipulation

What Is the Difference Between Legitimate Scientific Image Processing and AI Image Manipulation?

Scientific image processing can improve the visibility and analysis of authentic research data, while inappropriate AI manipulation may remove, introduce, or distort evidence. Learn how to distinguish defensible processing from misleading alteration using image provenance, scientific validity, and publication standards.

PART 0D · SUBPART 32 · GUIDE 473

AI for Scientific Image Enhancement

Can AI Be Used to Enhance Scientific or Research Images?

AI can enhance scientific images by reducing noise, improving visibility, and supporting image restoration. However, some techniques may suppress genuine signals or introduce details that were never observed. Learn how to distinguish defensible processing from misleading alteration, validate the results, and preserve the original evidence.

PART 0D · SUBPART 32 · GUIDE 472

AI-Generated Images in Research Papers

Can AI-Generated Images Be Used in a Research Paper?

AI-generated images may be acceptable in some research papers, but journal policies distinguish explanatory illustrations from data visualizations, primary scientific images, and graphical abstracts. Learn what publishers permit, what researchers must disclose, and how to avoid misrepresenting generated images as scientific evidence.

PART 0D · SUBPART 32 · GUIDE 471

AI-Generated Images for Research

Can Generative AI Be Used to Create Images for Research?

Generative AI can create illustrations, conceptual visuals, experimental stimuli, and synthetic images for research. However, researchers must distinguish generated content from observed evidence, evaluate its scientific accuracy, and consider disclosure, copyright, and research integrity requirements.

PART 0D · SUBPART 31 · GUIDE 470

Verifying AI-Assisted Mixed-Methods Integration

How Do Researchers Verify AI-Assisted Mixed-Methods Integration?

AI-assisted mixed-methods integration requires more than checking statistics and quotations. Learn how to verify relationships between quantitative and qualitative findings, audit joint displays, evaluate meta-inferences, and document the reasoning behind integrated conclusions.

PART 0D · SUBPART 31 · GUIDE 469

AI and Final Mixed-Methods Interpretation

Should Generative AI Produce the Final Integrated Interpretation in Mixed-Methods Research?

Generative AI can help draft the final integrated interpretation of a mixed-methods study, but it should not independently determine the research conclusions. Learn how to distinguish writing assistance from analytical judgment, evaluate AI-generated meta-inferences, and retain researcher accountability.

PART 0D · SUBPART 31 · GUIDE 467

AI and Overresolved Contradictions

Can Generative AI Overresolve Genuine Contradictions in Mixed-Methods Research?

Generative AI may turn genuinely contradictory quantitative and qualitative findings into a seemingly coherent explanation without sufficient evidence. Learn how to recognize overresolution, distinguish legitimate reconciliation from unsupported interpretation, and preserve meaningful uncertainty in mixed-methods research.

PART 0D · SUBPART 31 · GUIDE 466

AI for Explaining Conflicting Findings

Can Generative AI Explain Why Different Research Methods Produced Different Findings?

Generative AI can suggest why quantitative and qualitative research findings differ, including possible effects of measurement, sampling, timing, and context. However, plausible explanations are not necessarily supported explanations. Learn how to investigate disagreements systematically without allowing AI to invent causal mechanisms or conceal unresolved contradictions.

PART 0D · SUBPART 31 · GUIDE 465

AI for Mixed-Methods Meta-Inferences

Can Generative AI Generate Mixed-Methods Meta-Inferences?

Generative AI can propose meta-inferences by connecting quantitative and qualitative findings, but a plausible integrated conclusion is not necessarily a defensible one. Learn how to distinguish genuine integration from restatement, assess inferential quality, and verify AI-generated interpretations.

PART 0D · SUBPART 31 · GUIDE 462

AI for Mixed-Methods Joint Displays

Can Generative AI Build a Mixed-Methods Joint Display?

Generative AI can help researchers construct mixed-methods joint displays that connect quantitative findings with qualitative themes. Learn how to organize the evidence, distinguish integration from simple juxtaposition, and verify AI-generated interpretations.

PART 0D · SUBPART 31 · GUIDE 461

AI for Mixed-Methods Integration

Can Generative AI Help Integrate Quantitative and Qualitative Findings?

Generative AI can help researchers compare quantitative results with qualitative findings, organize integrated evidence, and explore possible interpretations. Its usefulness depends on preserving methodological context, examining contradictions, and independently verifying the conclusions it proposes.

PART 0D · SUBPART 30 · GUIDE 460

How to Report AI Use in Qualitative Analysis

How Should AI Use Be Reported in Qualitative Data Analysis?

Reporting that generative AI was used in qualitative analysis is not enough to explain its methodological contribution. Researchers should distinguish transcription, coding, theme development, and interpretation; document how outputs were verified; and disclose relevant data protection arrangements. Learn what to report, where to report it, and why transparent disclosure does not automatically establish methodological validity.

PART 0D · SUBPART 30 · GUIDE 459

Can AI Replace Qualitative Researchers?

Can Generative AI Replace the Interpretive Role of the Qualitative Researcher?

Generative AI can produce qualitative codes, themes, and interpretations, but generating findings is not necessarily equivalent to undertaking the researcher's interpretive role. Explore what human interpretation contributes, when AI can perform structured analytical tasks, and why methodological accountability cannot simply be delegated to a model.

PART 0D · SUBPART 30 · GUIDE 458

Does AI Coding Change Research Epistemology?

Does AI-Assisted Coding Change the Epistemological Assumptions of a Qualitative Study?

AI-assisted coding does not automatically change a qualitative study's epistemology, but how researchers use AI may introduce assumptions that conflict with their methodological commitments. Learn how automated classification, coding reliability, and delegated interpretation relate to interpretivist, constructivist, and other qualitative approaches, and how to evaluate methodological congruence.

PART 0D · SUBPART 30 · GUIDE 457

Can AI Have Positionality?

Can Generative AI Have Positionality in Qualitative Research?

Generative AI can describe researcher positionality and generate statements about identity, assumptions, and power. However, a model's technical characteristics and learned biases are not equivalent to the lived and institutional position of a qualitative researcher. Learn why this distinction matters, how AI-generated positionality statements can misrepresent research, and what researchers should document instead.

PART 0D · SUBPART 30 · GUIDE 456

Can AI Have Reflexivity?

Can Generative AI Have Reflexivity in Qualitative Research?

Generative AI can produce reflective statements, identify possible biases, and suggest questions about qualitative interpretation. But generating reflexive language is not necessarily the same as exercising methodological reflexivity. Learn why this distinction matters, what AI can contribute to reflexive practice, and why researchers must not confuse simulated reflection with their own analytical engagement.

PART 0D · SUBPART 30 · GUIDE 455

Can AI Perform Reflexive Thematic Analysis?

Can Generative AI Perform Reflexive Thematic Analysis if the Researcher Delegates Interpretation to It?

Generative AI can produce codes, themes, and analytical narratives resembling reflexive thematic analysis, but delegating interpretation raises a fundamental methodological problem. Examine why researcher subjectivity and reflexivity matter, how AI-generated analysis differs from researcher-developed meaning, and what the methodological literature says about substituting AI for the qualitative researcher.

PART 0D · SUBPART 30 · GUIDE 454

Can AI Understand Cultural Meaning?

Can Generative AI Understand Context, Irony, Ambiguity, and Culturally Specific Meaning in Qualitative Data?

Generative AI can recognize linguistic patterns associated with irony, ambiguity, and culturally specific expressions, but recognizing plausible meanings is not the same as establishing what participants intended. Learn how contextual misunderstandings can distort qualitative coding and interpretation, particularly in multilingual and cross-cultural research, and how researchers can evaluate AI-generated readings.

PART 0D · SUBPART 30 · GUIDE 453

Can AI Distinguish Description From Interpretation?

Can Generative AI Distinguish Description From Interpretation in Qualitative Analysis?

Generative AI can describe participant statements and propose interpretations, but it may blur the boundary between what participants actually said and what an analyst infers. Learn how to distinguish descriptive findings from interpretive claims, recognize unsupported assumptions, and evaluate whether AI-generated qualitative analysis offers genuine analytical insight.

PART 0D · SUBPART 30 · GUIDE 452

Can AI Write Analytical Memos?

Can Generative AI Write Qualitative Analytical Memos?

Generative AI can help draft qualitative analytical memos, organize emerging ideas, and suggest alternative interpretations. However, an AI-generated memo is not automatically a record of the researcher's analytical thinking. Learn how to use AI for memoing while preserving reflexivity, evidentiary grounding, and the development of original qualitative interpretations.

PART 0D · SUBPART 30 · GUIDE 451

Can AI Compare Themes Across Groups?

Can Generative AI Compare Qualitative Themes Across Participant Groups?

Generative AI can compare themes across participant groups and help organize similarities, differences, and contrasting experiences. However, apparent group differences may reflect sampling, interview conditions, or incomplete evidence rather than meaningful distinctions. Learn how to conduct evidence-linked comparisons without stereotyping participants or overstating qualitative findings.

PART 0D · SUBPART 30 · GUIDE 450

Can AI Identify Deviant Cases?

Can Generative AI Identify Contradictory or Deviant Cases in Qualitative Data?

Generative AI can help identify interview accounts that contradict emerging themes or challenge preliminary interpretations. However, unusual statements are not automatically deviant cases, and minority perspectives should not be dismissed as outliers. Learn how to use AI to examine disconfirming evidence, refine qualitative findings, and preserve contextual meaning.

PART 0D · SUBPART 30 · GUIDE 449

Can AI Apply a Qualitative Codebook?

Can Generative AI Apply a Qualitative Research Codebook?

Generative AI can apply predefined qualitative codes to interview transcripts and other textual data, but reliable classification depends on clear definitions and appropriate validation. Learn how to prepare a codebook for AI-assisted coding, evaluate ambiguous assignments, detect coding drift, and preserve methodological consistency.

PART 0D · SUBPART 30 · GUIDE 448

Can AI Develop Themes From Interviews?

Can Generative AI Develop Themes From Interview Transcripts?

Generative AI can identify patterns across interview transcripts and propose candidate themes, but recurring topics are not necessarily meaningful analytical themes. Learn how to assess theme quality, preserve participant context, examine contradictory accounts, and develop defensible interpretations from AI-assisted analysis.

PART 0D · SUBPART 30 · GUIDE 447

Can AI Generate Qualitative Codes?

Can Generative AI Generate Initial Qualitative Codes?

Generative AI can propose initial codes from interview transcripts, open-ended survey responses, and other qualitative data. However, plausible labels may overlook context, exaggerate meaning, or introduce assumptions. Learn how to generate useful coding suggestions, evaluate their quality, and preserve the researcher's interpretive role.