University of Eastern Finland Digital Marketing and Analytics Master's Thesis Guide: YK00DG59, 30 ECTS, DMA Seminar and Research

1. Start from the current 2026-2027 DMA thesis package

The University of Eastern Finland Digital Marketing and Analytics programme is a two-year, 120 ECTS Master of Science in Economics and Business Administration taught in English at Joensuu and Kuopio. For thesis planning, the controlling current study-plan object is Peppi programme 139667, code MDPDMA26. Its Advanced studies module contains four separate thesis-related objects that should not be merged: YK00DG59 Master thesis, Digital Marketing and Analytics, 30 ECTS; YK00EY95 Master Thesis Seminar, Digital Marketing and Analytics, 5 ECTS; YK00EY96 Research in Digital Marketing and Analytics, 5 ECTS; and YK00EJ61 Maturity Test, Master’s Degree in Business, 0 ECTS. The methodology module is another 15 ECTS. This structure matters because older study plans can show a different seminar package. Use your own current Peppi plan, not an older student’s screenshot or an archived curriculum.

2. The current structure is not the older 10 ECTS seminar model

A useful warning for DMA students is that historical programme pages can create a false impression. The 2024 structure showed a 10 ECTS master thesis seminar. The current 2026-2027 plan separates that work into YK00EY95 seminar 5 ECTS and YK00EY96 Research in Digital Marketing and Analytics 5 ECTS. The thesis itself remains a separate 30 ECTS course. This is not only a naming change. YK00EY96 has its own learning outcomes, participation expectations and assignments related to thesis development. Therefore, when you make your schedule or discuss missing credits, identify the exact current course code rather than saying only “the thesis seminar”. This prevents an old 10 ECTS rule from being copied into a current study plan.

3. YK00DG59 is the 30 ECTS thesis object

YK00DG59 is the current 30 ECTS Master thesis, Digital Marketing and Analytics. It is an English-language advanced-studies course in Business Economics and represents 810 hours of supervised independent work. The course expects the student to define and examine a well-bounded DMA-related research topic and to plan and complete the thesis project. Its learning outcomes connect research, data and information with decision-making in digital marketing and emphasise research that is analytical, methodologically reliable and practically valuable. The grade is 0-5. The course also states that submitted theses undergo authenticity verification through the university’s electronic plagiarism-detection system. Nothing in the current thesis course requires one universal statistical method, one software package or one data type.

4. YK00EY95 is a separate 5 ECTS Pass/Fail seminar

The current Master Thesis Seminar, Digital Marketing and Analytics is YK00EY95, 5 ECTS, English and Pass/Fail. It is separate from both the 30 ECTS thesis and the 5 ECTS YK00EY96 research course. The public course endpoint is unusually sparse and does not expose a current realization schedule, so students should not invent attendance dates, teachers or presentation counts from an older year. Check current Peppi teaching information and instructions given by the DMA team. The safe evidence boundary is the exact course identity, 5 ECTS scope, English language and Pass/Fail grading. Treat the seminar as part of the thesis-development environment, but use the current teaching implementation for cycle-sensitive participation requirements.

5. YK00EY96 is the DMA-specific bridge from topic idea to thesis execution

YK00EY96 Research in Digital Marketing and Analytics is another separate 5 ECTS Pass/Fail course for DMA master’s students. It is especially important because its content describes how the programme wants students to develop a thesis. The course covers the multidisciplinary nature of DMA research, the thesis process from ideation to submission, theoretical and methodological approaches in the field, academic and professional literature, key research databases and digital data sources, and the structure of the thesis from introduction to implications. It also includes developing the research problem and responsible use of AI and other technological tools. Current study methods require 80% presence in contact and online classes, DMA seminar participation activity, discussion and independent work. Assessment includes scientific-article analysis, evaluation of published DMA master’s theses and a seminar presentation.

6. The maturity test is the thesis summary, not an extra essay

YK00EJ61 Maturity Test, Master’s Degree in Business is 0 ECTS and Pass/Fail. The current course description states that the summary attached to the master’s thesis serves as the maturity test. It demonstrates familiarity with the thesis field and is assessed for content, structure and formal academic writing style. Both content and language must be approved. This makes the summary a formal degree component even though it carries no credits. Plan it as an accurate miniature of the final thesis: research problem, relevant context, design or data, central result and defensible conclusion. Do not introduce a claim in the summary that is stronger than the thesis evidence. Most importantly, UEF’s current AI guidance absolutely prohibits AI use in the abstract, including checking or translating it.

7. The methodology module is 15 ECTS and does not force one method

The 2026-2027 programme has a 15 ECTS Methodological studies module. YK00DO48 Research Methodologies in Business Studies, 5 ECTS, is compulsory. The remaining 10 ECTS are other methodological studies, with YK00DO49 Quantitative Research Methods in Business Studies and YK00DO50 Qualitative Research Methods in Business Studies listed as examples. This structure is important for thesis design. DMA is data-oriented, but the curriculum itself supports more than one defensible methodological route. YK00DO48 covers research problems, questions and designs; theory testing and theory building; quantitative designs including surveys, secondary or online data and experiments; qualitative designs including interviews, observation and secondary data; ethics; validity, reliability and trustworthiness; and AI in academic research. Method choice must follow the research question rather than the programme title.

8. A quantitative DMA thesis needs a design, not just a dataset

YK00DO49 demonstrates the type of quantitative competence available in the programme. It covers advanced quantitative analysis with software, statistical testing, regression and comparison of groups, with additional topics such as factor analysis, clustering, structural equation modelling and multigroup moderation. These are tools, not mandatory ingredients. A strong quantitative DMA thesis first states the unit of analysis, target population, sampling or data-generation process, outcome and explanatory variables, and the quantity the analysis is meant to estimate. Then it selects a model whose assumptions fit that design. Report effect size and uncertainty, not only p-values. If the data contain repeated customer, campaign, firm or platform observations, address dependence rather than treating every row as independent. Do not call an association causal unless the design supports a causal interpretation.

9. A qualitative DMA thesis is also a current programme-supported route

YK00DO50 confirms that qualitative research is part of the current Business Studies methods environment. It covers qualitative research questions and designs, case studies, grounded theory and narrative approaches, interviews, observations, focus groups, secondary materials and digital data, analysis and reporting, and management of AI and ethics. This is directly relevant to questions about consumer meaning, platform practices, organisational adoption, customer experience or marketing decision processes where numerical measurement alone may not answer the question. A qualitative thesis should still be systematic. Explain participant or document selection, access, data generation, recording and transcription where relevant, coding or interpretive procedure, reflexivity, saturation or information power where appropriate, negative cases, and how quotations or examples support the interpretation.

10. Mixed methods can be useful, but it is not a universal DMA requirement

Because YK00DO48 teaches both quantitative and qualitative designs, students may assume that a DMA thesis should combine them. The current sources do not establish such a requirement. Mixed methods are justified only when integration answers a research problem better than one approach alone. For example, platform analytics might show a behavioural pattern while interviews explain how users interpret the interface, or an experiment might estimate an effect while follow-up qualitative material explores mechanisms. If you combine methods, state the sequence, priority and integration point. Two disconnected datasets do not automatically make a strong mixed-methods thesis. A narrower single-method study can be more rigorous when it fits the question and can be completed within the thesis timeline.

11. Build the research problem around a decision-relevant gap

The programme describes DMA research as useful for understanding markets, consumers and digitalised business and as supporting digital-marketing decision-making. That does not mean every thesis needs a company client or immediate managerial recommendation. Start with a researchable problem: a theoretical inconsistency, unexplained behaviour, measurement problem, new digital context, platform feature, campaign mechanism or decision uncertainty. Then narrow the population, channel, platform, geography, time period and outcome. Avoid broad titles such as “The impact of social media on consumers”. A better question defines what social-media feature, which consumers, what outcome and what comparison or mechanism is being examined. YK00EY96 explicitly expects development of a research problem, so topic narrowing should happen early rather than after data collection.

12. Use theory to structure the study, not decorate the literature review

DMA topics can easily produce a long descriptive literature review about digitalisation. The thesis needs a tighter intellectual structure. Identify the concepts needed to explain the research problem and show how prior studies connect them. In explanatory quantitative work, theory should justify variables, expected directions, moderators or mediators before the results are known. In qualitative work, theory can provide sensitising concepts or an interpretive lens without forcing observations into predetermined categories. In design-oriented or practice-facing work, theory should still explain why the chosen problem matters and what knowledge the thesis adds. YK00EY96 asks students to identify the logic and structure of academic and professional literature and integrate theoretical and methodological insights into thesis work. Use that as a standard for coherence.

13. Digital traces and platform data need provenance documentation

DMA students may work with web analytics, platform exports, advertising data, CRM records, e-commerce logs, app events, reviews, social-media material or commercial databases. These sources can look objective because they are automatically generated, but the variables are produced by systems with their own definitions. Record where each field came from, the extraction date, coverage period, filters, account settings, attribution rules, missing events, bot filtering, identifier changes and any transformation you make. Platform metrics may be redefined over time. An API response can change after a platform update. A campaign dashboard may use an attribution model that differs from your research construct. Preserve a data dictionary and transformation log so another reader can understand the analytical dataset.

14. Publicly visible digital data are not automatically free of ethics or legal issues

A common mistake is to treat any publicly viewable post, review or profile as unrestricted research data. UEF’s data-management and ethics guidance requires case-specific attention to personal data, participant expectations, contracts, access rights and safe processing. Platform terms of service, API conditions, copyright and privacy can matter even when a webpage is visible without login. If you combine identifiers across sources, re-identification risk can increase. If you quote searchable user text, anonymity can fail even after removing a username. Decide what you actually need to retain, whether identifiers are necessary, how data will be stored, and whether publication of examples could expose a person or organisation. Discuss uncertain cases with the supervisor and appropriate UEF support before collection.

15. Surveys and experiments need measurement and implementation controls

DMA questions often involve attitudes, intentions, customer experience, trust, engagement, adoption or response to digital stimuli. A survey should define constructs before collecting responses and distinguish validated scales from researcher-created items. Document translation, pretesting, attention checks, response-quality rules, missing-data handling and exclusions. An experiment should specify treatment, comparison condition, randomisation, manipulation checks, outcome timing and exclusions before looking for a preferred result. Online recruitment can create selection bias, and repeated exposure can weaken independence. If the thesis tests a mediation or moderation model, show that the design and sample support the model rather than adding complexity only because the software can estimate it.

16. Interviews and qualitative digital research require a transparent chain from data to interpretation

For interviews, explain who was eligible, how participants were recruited, why the sample fits the question and how consent and recording were managed. An interview guide should connect to the research problem but remain open enough for unexpected insights. For digital ethnography, online communities or user-generated content, define the field boundary and observation period. During analysis, document how codes or themes were developed, revised and connected to the final claims. Quotes should illustrate an interpretation rather than replace analysis. If AI-assisted transcription or coding is considered, check current UEF rules, confidentiality and data-protection implications first and disclose permitted use. The student remains responsible for the analytical decisions and accuracy of the record.

17. Measurement quality is central when DMA constructs are indirect

Many marketing constructs cannot be observed directly. Loyalty, trust, engagement, perceived value or digital capability may be represented by multiple survey items or behavioural proxies. State how each construct is operationalised and why the measure matches the concept. For multi-item scales, examine reliability and construct validity appropriately. For behavioural proxies, explain what the metric can and cannot represent. Click-through rate is not automatically persuasion; dwell time is not automatically attention; sentiment score is not automatically attitude. YK00DO49 includes construct validation and bias assessment, while YK00DO48 requires attention to validity and reliability. Treat these as reasoning tasks, not as a checklist after the main analysis.

18. Separate prediction, explanation and causal claims

Analytics software can produce accurate predictions without explaining why an outcome occurs. An explanatory model can estimate associations without supporting intervention claims. A causal design asks a different question again. State the purpose before choosing the model. If the thesis predicts churn, conversion or response, report validation strategy and performance on data not used to fit the model where feasible. If the goal is explanation, discuss theoretical interpretation and uncertainty. If the goal is causal inference, justify identification assumptions, treatment assignment or quasi-experimental strategy. Avoid converting “X predicts Y” into “X causes Y” in the discussion. DMA’s practical value increases when the claim is calibrated to what the design actually establishes.

19. Evaluate validity, reliability and trustworthiness throughout the process

Quality control is not one paragraph at the end. In quantitative work, consider construct validity, internal validity, external validity, reliability, missingness, measurement error and model assumptions. In qualitative work, consider credibility, transparency, reflexivity, dependence on context and whether the evidence supports competing interpretations. In secondary digital data, provenance and platform measurement become part of validity. YK00DO48 explicitly covers validity, reliability and qualitative trustworthiness. Use the concepts that fit your design rather than mechanically listing all of them. A limitation section is stronger when each limitation explains the direction or scope of the uncertainty and how it affects the conclusion.

20. Central ethics review is not automatic for every master’s thesis

UEF’s Research Ethics Committee states that it does not review individual master’s theses as a rule because master’s-level students should not normally conduct designs requiring prior ethical review. A thesis that is part of a broader research project can be covered by that project’s ethical-review process. This boundary must not be simplified into “ethics approval is never needed”. Research ethics, informed participation, privacy, data protection, power relations, risk and responsible reporting still apply. Public or registry data can also raise issues when personal data are combined in ways that threaten privacy. Resolve permissions and ethical questions before collecting data, particularly where participants, sensitive information, tracking data or vulnerable groups are involved.

21. Make a data-management plan before the dataset becomes difficult to control

UEF Data Support provides student resources specifically for master’s-thesis data management. Before collection, decide what data will exist, who has access, where it will be stored, how identifiers will be separated, what will be backed up, how versions will be named, and what will happen after the thesis. Interview and survey data almost always contain personal data in some form. Marketing datasets may also contain customer IDs, IP addresses, device identifiers, purchase history or location-related variables. Do not upload sensitive or confidential data to consumer cloud or AI tools without an approved basis. Document the analytical dataset and code so the results can be checked while respecting access restrictions.

22. AI can support permitted parts of the thesis, but the abstract is an absolute exception

UEF currently permits AI support in theses when use is transparent, responsible and within applicable instructions. The student remains responsible for accuracy, references, analysis and final text. Report the tool, purpose and timing of material AI use. A course, unit or supervisor may impose stricter limits. The hard boundary is the thesis abstract: UEF states that AI must not be used at all in the abstract, including checking or translating it, because the abstract often functions as the maturity test. This is directly relevant to DMA, where students may otherwise be tempted to use generative AI or language tools as part of a digital workflow. Draft and revise the maturity-test summary yourself.

23. Company collaboration does not make the assessed thesis confidential

DMA projects can involve companies, campaigns, customer data or platform partners. The university’s advanced-studies thesis is public as a rule, and UEF recommends Open Access publication. Design collaboration so commercially sensitive information does not need to appear in the assessed public thesis. Agree early on data access, permitted aggregation, anonymisation or pseudonymisation, review of factual confidential details and what can be published. A company can flag trade secrets or factual risks, but the academic question, analysis and conclusion remain part of the student’s independent work. If a separate confidential deliverable is needed, keep it distinct from the university thesis and confirm the arrangement with the supervisor.

24. Supervision and seminar work should be used as research-control points

The 30 ECTS thesis is supervised independent study, while YK00EY95 and YK00EY96 create structured points for feedback and participation. Use these points deliberately. Early supervision should resolve the research problem, access feasibility and method-data fit. Before collection, confirm ethics, data protection and measurement. Before analysis, freeze key inclusion/exclusion and coding decisions where feasible. Before final drafting, check whether the results actually answer the research question. YK00EY96’s article analysis, evaluation of published DMA theses and seminar presentation can be used to calibrate quality against real examples. Keep a decision log after meetings so revisions are traceable and repeated discussions do not consume the thesis schedule.

25. A defensible thesis structure makes the reasoning visible

YK00EY96 explicitly covers the main components of introduction, literature review, methodology, analysis/findings, discussion and implications. The introduction should define the problem, gap, question and contribution. The literature review should build the conceptual logic. The methodology should explain design, data, measurement or interpretation, ethics and analysis clearly enough to evaluate reliability. Findings should report evidence without hiding contradictory results. Discussion should return to the theory and question rather than repeating tables. Implications should be proportional to the evidence. Conclusion should state what the study establishes and where uncertainty remains. Appendices can hold instruments, extended diagnostics or coding materials, but essential reasoning should remain in the main thesis.

26. Do not turn the faculty’s indicative page range into a hard DMA rule

The current faculty study guide describes a general indicative pro gradu length of about 60-100 pages and notes that scope can vary by discipline. The DMA thesis course itself does not establish a universal page or word limit in the public record used for this guide. Therefore, do not treat 60 pages as an automatic minimum or 100 pages as a universal maximum. Follow the current DMA/Business School instructions and supervisor guidance for formatting and expected scope. Research quality depends more on bounded questions, transparent method and evidence than on reaching a page count. If a local template or seminar instruction gives a more specific limit, that lower-level current instruction should control your submission.

27. Use the current Business School/DMA assessment criteria, not a substituted generic rubric

YK00DG59 establishes the 0-5 scale and says evaluation follows faculty guidelines. The faculty guide provides general goals and allows more specific unit requirements. UEF’s general thesis page also directs students to the faculty or subject curriculum for assessment criteria. This means a student should obtain the current Business School/DMA rubric or seminar assessment instructions when preparing the final thesis. Do not replace a programme-level criterion with an old faculty table, another Business School programme’s rubric or a PT Writers interpretation. The safe practical strategy is to map your current rubric against the research problem, theoretical grounding, method justification, evidence quality, analysis, discussion, scientific communication and independence, and ask the supervisor to clarify any criterion that is not accessible publicly.

28. Turnitin is a formal pre-submission gate

UEF’s current submission workflow requires the completed thesis to be submitted to Turnitin in eLearn Moodle before official review submission. The supervisor or principal supervisor reviews the plagiarism-detection report and gives permission to proceed. YK00DG59 independently states that submitted DMA theses undergo electronic authenticity verification. Treat Turnitin as an authorship and source-use control rather than a target similarity percentage. Resolve quotation, close paraphrase, missing citation, copied method descriptions and reuse of your own prior text before the final check. If the report reveals a genuine problem, correct the scholarly practice rather than trying to manipulate the similarity score.

29. Official submission is one PDF/A through UEFe-Services

After the plagiarism check is approved, convert the thesis and required appendices into a single PDF/A file and submit it through UEFe-Services. Complete the metadata carefully because details such as unit, supervisor, thesis title and publication choice move into the formal process. Check the exact title rather than copying a version with formatting errors. The submission form also asks for the selected public-access level, and UEF recommends Open Access. Ensure figures, tables, links and document structure are accessible. Do not wait until submission day to convert to PDF/A because conversion can expose font, image, bookmark or accessibility problems that take time to fix.

30. Two examiners, a written statement and your response are part of the formal assessment

UEF Education Regulations require two examiners for an advanced-studies thesis, and as a rule one should be the supervisor. The supervisor proposes reviewers and a decision-maker; the head of department or unit approves the reviewers. The reviewers provide a written statement and grade proposal within one month of appointment, subject to excluded no-teaching periods. Before the final decision, the student has an opportunity to respond. In the current online process you may accept the proposal, submit a response or request one suspension of review and revise before resubmission. If proposals differ, or the grade proposal is not accepted, a third reviewer may be appointed. A proposed grade is therefore not itself the final grading decision.

31. Publication planning starts before the final upload

After grading, the administrator records the thesis in Peppi and the library processes it according to the access level selected during submission. If Open Access is selected, the thesis can be browsed, copied and printed online. The access level can later be changed by written request. Because the thesis is public as a rule, privacy and company-confidentiality decisions should be solved much earlier than this stage. If raw data cannot be opened, that does not necessarily prevent an academically transparent thesis. Explain data restrictions, processing and analytical procedures at an appropriate level while withholding protected material. Data publication and thesis publication are separate decisions and should not be confused.

32. Build the timeline backwards from research and graduation gates

For semester-end planning, UEF currently advises submitting a master’s thesis for review by 30 April for spring graduation planning and by 31 October for autumn, with degree-certificate applications by 31 May and 30 November respectively. These are university planning dates, not replacements for DMA seminar, YK00EY96 or supervisor deadlines. Work backwards through topic approval, YK00EY96 research development, literature and theory, method selection, ethics and data-management decisions, access or recruitment, data cleaning or transcription, analysis, seminar presentation, full-draft feedback, abstract written without AI, Turnitin, PDF/A conversion, UEFe-Services submission, examination and possible revision. Before submission, recheck your current Peppi study plan and the current Business School/DMA instructions because teaching implementations and local guidance can change between academic years.