Company data in a master's dissertation - how to prepare your material for the research chapter

You have a file from a placement or a report from work, and you can see how it might fit into your dissertation. Before you start running the numbers, establish two things. First, which information can you use and show in your dissertation? Second, what question can you answer with that information? If the two fit together, you can carry out your own analysis of company data and describe it in your research chapter.

Start with a note. You can do this before sharing the spreadsheet. Write down one research question, the name of the source, the period it covers and what one row represents. Below that, list the fields you need to answer the question and note who can confirm how you may use them. For instance, “Did the proportion of orders delivered on time differ between May and June?” calls for three pieces of information: which orders belong to each month, when they were due and what receipt date the delivery confirmation records. Take that note to your supervisor and the person responsible for the data at the company. The steps below follow the material from the original export through the agreement on what can be used, then show a worked result and a sample passage from the chapter.

Does access at work or on a placement mean you can use the file?

Access to a system through your role lets you carry out certain tasks for the company. Using material for your dissertation, presenting a result at your defence and sending a file to someone for advice are further uses, and each may be subject to different conditions. Check the company's rules and ask who can decide what you may use. Your team manager may know what each column means, while someone else is responsible for authorising disclosure. Ask about both before you build your project around the full export.

At your university in Poland, check whether your supervisor accepts the source, what your account of the material needs to include and how you can present the result in your dissertation, at your defence and in any appendices. If the material comes with particular conditions, find out whether and how the university can meet them. A printed copy, a file sent to your supervisor, a file submitted through the university system and slides for your defence are all places where the information could appear. Note which version of the result can go in each. Before including information supplied under specific conditions, check that you can meet the requirements of both the company and the university.

What should you agree with the company before requesting data?

Show the company a short description of your project before requesting the whole register. Give the period, the part of the business you want to study, the type of record, the fields you need and the form of the result. For delivery performance, you may only need the number of orders covered by the analysis, the number with a known outcome and the number delivered on time. If you want to compare reasons for delays, you also need comparable categories and the number of cases for which a reason was recorded. Ask for more detail when you can show how it answers your question.

Separate four decisions when you speak to the company.

• Access to the material. Find out whether the company will share a file, prepare a summary or let you carry out agreed calculations in its own systems.
• Scope of the analysis. Explain what you need the material for, which fields, periods and groups you will cover and whether you intend to combine it with another source.
• What can be disclosed. Agree whether you may name the company, state the period, show counts and results or include breakdowns, extracts from records or screenshots.
• Who can access it afterwards. Agree who can see the file, in what form, for how long and what happens to the copy when the project ends. Consider sharing it with your supervisor separately from sharing it with an external adviser.

Confirm the terms in a form accepted by both the company and your university. Keep a note of what was agreed and when, so you can refer back to it if your question changes or someone asks for another chart. In your methods section, you can later describe how the material was supplied and what you were authorised to do with it, using the agreed wording. Share the email, signed document or request itself only through the process your university requires and within the terms you agreed.

How can you write to the company before you have the material?

Explain what result you want to produce and give the company the option of preparing a smaller dataset. A request framed this way makes clear that you can compare months without looking at delivery addresses or customer names. If the company prefers to produce a summary, ask whether it can also provide definitions of the fields and a brief note on how it made the calculations. You can then explain where the figures came from and which assumptions were used.

I'm preparing a master's dissertation on delivery performance. I'd like to compare the proportion of orders delivered on time in May and June for an agreed category of orders. Could the company provide monthly counts of the orders covered, those delivered on time, those delivered late and those whose outcome could not be established? I would also need to know how the delivery deadline and confirmation are defined, and how exclusions are handled. Could you tell me which information I may include in the dissertation I submit to the university and present at my defence? If further agreement is needed, please let me know who I should speak to.

If the working file is already on a company device, start by describing the fields and asking how the material can be used. You can send a message like the one above without attaching the file. If the reply is “you can use the report”, ask whether that means analysing it within the company, including results in your dissertation or sharing a copy with your supervisor. If the information you need cannot be agreed, reduce the scope of the project and show your supervisor what material will actually be available.

What should you check about personal data and confidential information?

One column may be needed for the calculation, while another may simply reveal who placed an order. Treat those fields separately. A name, email address or telephone number is rarely needed to compare monthly delivery performance. An order number may help the company find duplicates in its own system. Once it has done that, the number need not appear in the summary for the dissertation. Consider commercial information separately. Even without names, a spreadsheet may reveal prices, contractual terms, routes or results for a particular customer.

Article 11 of Poland's Act on Combating Unfair Competition sets out when technical, organisational or other commercially valuable information may constitute a trade secret. The assessment includes how readily the information can be obtained and what has been done to keep it confidential. Ask an authorised person at the company about the status of your specific material. Discuss each type of data, including summaries that could reveal details when combined with other information. Check the company's confidentiality terms and the rules on handling personal data separately, because both may apply to the same dataset. If your project requires personal data, show your plan to the person responsible for data protection at the company and agree with the university who will review how it is to be processed.

How is removing a name different from anonymising the data?

Replacing a name with a code can be part of pseudonymisation. Under the EU GDPR, the additional information needed to link the data to a person, such as the key linking the code to the name, must be kept separately and appropriately protected. The information may still be personal data. When assessing whether material is anonymous, the question is whether a person could still be identified by means reasonably likely to be used. Consider how different pieces of information could be combined as well as what any single field reveals. Article 4(5) of the EU GDPR defines pseudonymisation and recital 26 explains how to assess identifiability. Poland's Personal Data Protection Office (Urząd Ochrony Danych Osobowych) uses the example of published documents to show how other features can still make a person identifiable after their name has been removed.

In practice, exact dates, locations, job roles or comments can remain in a spreadsheet after names have been deleted. Imagine a report of a breakdown on a shift when only one person was working in a particular role. The summary might still make that person identifiable. In a delivery project, a similar issue can arise with a customer: an unusual route, a date and the value of the goods may reveal which order a row refers to. The first example concerns identifying a person. The second may also disclose commercial information. The appropriate approach depends on the dataset and the terms agreed for the intended use.

Are summary figures always safe to share?

Aggregation means combining individual records into larger groups, for example reporting a monthly delivery count instead of listing each delivery. This often reduces the need for detailed information. A summary can still expose an individual case if it reports figures for very small groups. If the company has approved a monthly total and your chart breaks it down by customer or by a rarely used service, check that breakdown before using it. The same applies if two charts look general on their own but subtracting one from the other reveals a small group.

Start with the least detailed information that can answer your question. For an overall comparison of delivery performance in two months, monthly totals are enough if you have the necessary numerators and denominators and know how they were calculated. To compare locations, you need figures for each location. If those figures cannot be presented safely, agree a question you can answer with the material you can document. State how the data were grouped in your chapter and make sure the scope of your conclusions matches that level of detail.

When should you check the university's procedure for confidential dissertations?

Check it when your project would use information the company will not let you disclose through the university's standard submission or defence process. Poznan University of Technology (Politechnika Poznańska) tells students about a confidentiality agreement and directs them to its Placement and Careers Centre (Centrum Praktyk i Karier). The University College of Enterprise and Administration in Lublin (Wyższa Szkoła Przedsiębiorczości i Administracji) describes a request and decision process for treating all or part of a dissertation as confidential. These are the procedures at those institutions. Check the requirements of your own department, including who submits the documents, when they are due and how to prepare the copies. Clarify the arrangements before choosing your material, as they can affect how you present results for assessment.

Ask three questions about your project. Who can receive the full material? What must go into the submitted version and the presentation for your defence? Once you limit the amount of information disclosed, can you still show how you reached the result? If the last answer is unclear, speak to your supervisor about using an available summary or changing your source. The chapter needs to show enough of your reasoning for it to be assessed, even if the underlying details stay with the company.

How can you tell whether a company report answers your question?

A system export, a monthly report and an order-handling procedure can all be sources of existing data. The original records are usually created for a purpose of the company's own, such as reconciling transactions or monitoring orders. Even if the company produced the export at your request, it can contain data that were collected earlier. Marta Makowska and Rafał Boguszewski (2013) discuss several types of existing data and how reliable they are. Marcin Rechciński, Jarosław Balon and Małgorzata Grodzińska-Jurczak (2017) use their own project to show how to assess the usefulness of sources against different research aims. You can apply the same idea to your dissertation by asking what each source can tell you about a particular question.

Establish what one row represents and how the record was created

Depending on the system, one row could represent an order, a parcel, a line item or a change of status. If an order has three parcels, treating the three rows as three orders will distort the denominator. Similarly, a “closed” date might record physical receipt or simply the moment an employee closed the case in the system. Ask for field definitions and trace a few cases within the terms on which you can work with the material. Note which event represents the concept you are studying and which is an administrative step.

If the company provides a summary only, ask how it was produced. “Completed orders” may refer to orders marked as delivered, or it may include every closed case, including cancellations. “On-time delivery” may be measured against a date agreed with the customer or a date entered in an operational plan. If you cannot establish what the indicator means, discuss that gap with your supervisor. Changing the question to a description of the company's indicator is useful only if you know enough about what it covers.

Check the period, completeness and comparability

Decide whether you are comparing orders placed, orders due for delivery or orders actually delivered in each month. An order due on 31 May and delivered on 2 June may belong in the May group if the scheduled date determines the month. If you select June orders by their actual delivery date, you will be comparing two differently defined groups. Use the same rule for both periods. Record the date you stopped observing the orders so you can distinguish missing information from a status that could still have changed.

Check that the register covered the same operations in both months and find out whether the system or the requirement to enter a status changed. Look for orders moved between periods, amendments, duplicate records and unusual gaps. If one branch used a different system for a week, a whole category of orders may be absent from the file you have. The percentage calculated from the remaining records would then describe the available material and would give an incomplete picture of the company's operations that month.

Apply a similar check to non-numerical data. A complaints procedure tells you which steps the company expects staff to follow. The case history tells you which actions were entered in the system. To find out how customers felt about the contact they received, you would need a different source. If you use a document alongside a register, explain what each source does and how you will compare them. Be clear whether you are comparing the expected procedure with a record of what happened or bringing together two accounts of the same process.

A logistics example. From an order export to a result in the dissertation

In this fictional example, Ola is writing a master's dissertation on on-time delivery. The company description, message, figures and chapter extracts were written for this guide. In a real dissertation, you need to show where the data came from and what terms apply to their use.

What does Ola have at the start?

The company uses an order management system. Staff can see the order number, the customer, the address, the deadline agreed with the customer, the current status and the delivery confirmation date. The system also records some changes in status. An internal export for two months contains records for orders in one service category. Ola wants to find out what proportion of orders with a known outcome were delivered on time in May and June 2026. She originally wanted to compare customers and describe the reasons for delays as well. After speaking to the company, she learns that she will not get a breakdown by customer, and that reasons were entered as free-text comments and were not available for every order. She drops those two questions before running the calculations.

Her note to her supervisor now sets out the scope: order delivery performance for one service category over two months. Her method is to compare the percentages descriptively using a company summary. Ola also proposes reporting the proportion of orders whose on-time delivery status could not be established. She and her supervisor agree that the project is feasible if the company confirms the definitions used in the summary and allows the results to be included in the dissertation. The analysis will describe the differences she observes. Studying the reasons would call for other material and a different research design.

What does the company share and what stays in its system?

In the example, the person authorised to share the material agrees to give Ola two monthly summaries without customer names, order identifiers or individual delivery dates. Ola may include a general description of the process, the period and the approved counts and percentages in her dissertation and at her defence. The company prepares the summaries internally using agreed rules and explains how it identified duplicates and which cases it excluded. Ola uses that information in her account of the procedure. She can say that she reviewed the summaries and the explanation, and should make clear that she did not check each order in the system herself.

Ola needs four figures for each month: the number of orders covered by the analysis, the number delivered on time, the number delivered late and the number with an unknown outcome. She also asks for totals showing how the number of rows in the export became the number of orders in the analysis. This helps her check the arithmetic and explain to the reader how the denominator was obtained. The customer field serves neither purpose. She also has no need for order values, employee names or comments.

How did she define the deadline, delivery and an unknown outcome?

In this example, the month is determined by the delivery date agreed with the customer, taken from a specified version of the schedule. An order is on time if it is delivered by the end of that day. The system's delivery confirmation field records the date the order was received, which can differ from the day an employee entered the information. In logistics, Agnieszka Gaschi-Uciecha (2025) also refers to an agreed schedule when discussing supplier reliability. The specific calculation rules in Ola's example follow from her research design and how the company made its records. The same observation cut-off applies to May and June, 31 July 2026. The analysis uses the records available by the end of that day.

The company and Ola agree three rules. Duplicate entries for the same order are removed before counting. Orders cancelled before the agreed delivery date are excluded because the orders were no longer due for delivery when the deadline arrived. For the remaining orders, a missing confirmed date alone does not establish that a delivery was late. The outcome is recorded as unknown if there is no other reliable record that resolves its status. If the register reliably shows that an order was still awaiting delivery at the cut-off, record this separately from missing information. In a real study, also account for any orders cancelled after the deadline if they appear in the material. The company identified neither of these additional cases in the fictional example.

Here are three made-up cases.

• An order was due on 18 May and the confirmation shows it was received on 18 May. It belongs in the May group and was on time.
• An order was due on 31 May and the confirmation shows it was received on 2 June. Under the selection rule, it belongs in the May group and was late.
• An order was due on 9 June. It is in the agreed set, but the receipt date cannot be established from the confirmation records. It belongs in the June group and its outcome remains unknown.

If a delivery was rescheduled, the study needs to say which date is the reference point. The original date tells you whether the initial commitment was met. A later date can tell you whether a revised commitment was met, provided you can establish that it was agreed. Choosing a different version of the date for each order after seeing the outcome takes away the common basis for comparison. In this example, Ola uses one documented version of the schedule and applies the same rules in both months.

How does Ola turn rows into order counts and percentages?

May. The export had 186 rows. Removing 3 duplicate entries and excluding 3 orders cancelled before their due dates left 180 orders. Of those, 150 were delivered on time, 18 were delivered late and 12 had an unknown outcome. The outcome could therefore be established for 168 orders.

June. The export had 227 rows. Removing 4 duplicate entries and excluding 3 orders cancelled before their due dates left 220 orders. Of those, 176 were delivered on time, 24 were delivered late and the outcome was unknown for 20. There were 200 orders with a known outcome to include in the calculation.

Ola divides the number of on-time deliveries by the number of orders with a known outcome and multiplies by 100. In May, 150/168 gives 89.3%. In June, 176/200 gives 88.0%. The observed difference is 1.3 percentage points after rounding. Separately, she reports the share of unknown outcomes among all orders covered: 12/180, or 6.7%, in May and 20/220, or 9.1%, in June. Readers can see that the two percentages use different denominators. The first describes delivery performance among orders with a known outcome. The second shows what share of all orders covered could not be assessed.

Across the two months, 326 of the 368 orders with a known outcome were delivered on time, giving 88.6%. Ola calculates that percentage from the combined counts. A simple average of the monthly percentages would give the two months equal weight, although they include different numbers of orders. The right choice depends on the question. To find the proportion of on-time orders in the whole dataset, calculate it across orders. If your question is about the average monthly result, the unit is the month and you need to explain why you chose that measure.

What if every order with an unknown outcome had arrived on time or late?

Ola can show what happens under two extreme assumptions. If every order with an unknown outcome had been late, the on-time share across all orders covered would be 150/180, or 83.3%, in May and 176/220, or 80.0%, in June. If they had all been on time, the shares would be 162/180, or 90.0%, and 196/220, or 89.1%, respectively. The counts in the example set those bounds. They help you judge how much the missing outcomes could matter compared with the difference between the two months.

The ranges of 83.3% to 90.0% and 80.0% to 89.1% overlap. Under some assumptions about the unknown outcomes, May would have a higher on-time share. Under others, June could come out ahead. The available information does not establish what happened. Ola therefore reports the difference among orders with a known outcome and shows the limits of what she knows about all orders. If the company can check why statuses are missing, it is worth explaining whether the gap is linked to a particular way of recording deliveries, for instance. If the terms of access mean that this information is unavailable, state that in the chapter and explain what it means for the conclusion.

How do you describe the source, calculations and results in your dissertation?

In your research chapter, give separate accounts of access, how the material was prepared and what you analysed yourself. Readers should be able to see who prepared the summary and what decisions they made. The company might have removed duplicates and added up records. The student might have agreed the definitions, checked that the counts were consistent, calculated percentages and compared the months. If you received the data with the indicators already calculated, say so and show how you worked with the results. Your contribution could be choosing the criteria for comparison, combining two approved sources or assessing the effect of incomplete records. Explaining how you obtained the material helps readers see which of these steps you carried out.

What goes into the methods section if the company supplied summary data?

State the research question, the process and period covered, the unit of analysis and the inclusion rules. Explain how the report was produced, what it contains and what you could not check. Define on-time delivery and say how duplicates, cancellations and unknown outcomes were handled. Explain which counts form your denominators. Say whether you worked with individual records or monthly totals provided by the company. When you describe the terms of access, keep within what you are allowed to share. You may be able to name the type of register and the department without naming an employee or attaching internal emails.

Here is how part of the methods section might read in Ola's fictional project.

The material consisted of summaries prepared by the company from its internal order register. The summaries covered an agreed service category with orders due in May or June 2026. The unit of analysis was the order. Each order was assigned to a month using the delivery date agreed with the customer, taken from a specified version of the schedule. An order counted as on time if the recorded receipt date was no later than the agreed date. The cut-off for both months was 31 July 2026. The analysis used records available by the end of that day. The company removed duplicates from the summaries and excluded orders cancelled before their due date. Cases with insufficient information to determine the delivery status were recorded as having an unknown outcome. The analysis calculated the share of on-time orders among those with a known outcome and the share of unknown outcomes among all orders covered. The study used the monthly counts and the description of how they had been prepared.

In your dissertation, include enough information to identify the version of the source, such as the agreed name of the summary and the date it was prepared. If the company updated the register after the initial export, state which extract you used. If you use two files, explain how they were linked or why you present them alongside each other. Otherwise, readers may apply a rule described for one dataset to another. Agree with your supervisor how to identify an internal source under your programme's referencing requirements.

How can you make clear what a percentage refers to?

Start with the counts. State which orders the percentage is based on, then give the percentage and the difference. Finally, identify the unknown outcomes and explain the limits of the interpretation. If you have space for a chart, make its denominator clear and mention missing outcomes in the caption or the adjacent paragraph. “On-time delivery in May and June” is too broad a caption if the chart only includes orders with a known outcome.

Here is a sample results paragraph.

In May, 150 of the 168 orders with a known outcome were delivered on time (89.3%). In June, 176 of the 200 orders with a known outcome were delivered on time (88.0%). The on-time share among orders with a known outcome was 1.3 percentage points lower in June. The outcome could not be established for 12 of the 180 orders covered in May (6.7%) and 20 of the 220 covered in June (9.1%). The difference of 1.3 percentage points applies to orders with a known outcome. Depending on the outcomes of the remaining orders, the full dataset could show a higher percentage in June.

In the discussion, consider what could limit the comparison. May and June may have differed in the number or mix of orders, how staff recorded them or how complete the delivery confirmations were. Say which differences the material establishes and which would need to be checked. This design lets you describe the result for the part of the operation you studied. To say what caused the difference, you would need data on that cause and a way to separate it from other changes. Two monthly results may justify looking into the process further. Identifying a longer-term trend depends on the period covered and whether later figures are comparable.

What can you write if the company only shared a percentage?

Start by asking for the definition and the denominator. If the report says “91% on time”, ask which orders that percentage covers and whether it includes cancellations, unknown outcomes or only orders for one service. Also ask whether every month was calculated in the same way. If the company can provide the numerator (the number of orders delivered on time) and the denominator (the number of orders the percentage is based on), recalculate the percentage. If you cannot obtain the denominator or check the exclusions, describe the figure as an indicator reported by the company. State what you could not verify and ask your supervisor whether that account meets the aim of the dissertation.

Percentages without counts cannot tell you the overall share of on-time orders across two months. If 90% in one month refers to 10 orders and 80% in the next refers to 100, taking 85% would give the average of the two monthly percentages. The share of on-time orders across the whole dataset would be 89/110, or about 80.9%. You need the counts to reach that conclusion. Ask which unit should have equal weight: the order or the month. If the question is about orders, show the basis of the calculation in your dissertation or limit the conclusion to the percentages in the reports.

What if the company will not let you show the results?

Go back to what you agreed. Sometimes the company can supply less detailed figures or a summary for a longer period. Check whether you will still have enough information to answer the question. Combined figures for two months can describe on-time delivery across that period. To compare May with June, you need separate figures. If the company can provide only a general description of the process, you could study the content of the procedure if it fits your programme and question. Changing the material changes what the dissertation can establish.

If confidentiality makes it impossible to show how you reached the result, even under your university's confidential submission procedure, discuss another source with your supervisor. Check whether you need to amend the approved topic or scope. Comparing the designs before and after the change will help you see which parts of the chapter are still useful. The explanation of logistics concepts may still provide context. The definition of an indicator that requires the company's own register may no longer be useful. In the revised plan, set out how the new source changes the question, period, unit and method.

How can you use a public source if the company data fall through?

If her supervisor agrees to the revised question, Ola could design a new project using public freight transport data. In its 2 June 2026 release Przewozy ładunków i pasażerów w 2025 r. (Freight and passenger transport in 2025), Statistics Poland (Główny Urząd Statystyczny) provides an overview and tables showing transport volumes by mode. You could compare freight volumes for a particular mode of transport across periods. The figures describe that mode of transport in each period. They cannot tell you whether orders at Ola's company arrived on time. In the dissertation, identify the tables used and check their units and accompanying notes.

Public data can also provide context for a company study if the two sources serve distinct purposes. You could describe the national picture for freight transport and then examine on-time delivery in one company process. To compare figures directly, you need matching indicators, units and periods. Statistics Poland measures something different from the company's on-time delivery figures. Its data cannot fill gaps in the company register, and a national average for another indicator cannot tell the company what result it should aim for.

Can you use fictional figures in a master's dissertation?

Fictional figures can be used in an exercise or a clearly labelled demonstration of a calculation if it serves a purpose in the dissertation. Ola's example shows how to identify the denominator and handle an unknown status. When you study a real company and ask about its performance, conclusions must be based on actual data you are allowed to use. If you cannot obtain those data, agree a change of material, question or project design with your supervisor. Readers can then see which figures come from the study and which illustrate the method.

What should you check before sending the chapter to your supervisor?

Read through the chapter from the question to the conclusion. At the question, note which material and period can provide an answer. In the methods section, check that someone can reconstruct the inclusion rules, the definition of the outcome and the treatment of missing information. Give the numerator and denominator for each important percentage. Where a figure came from the company, say how it was obtained. At the conclusion, check whether it refers to the dataset, the whole company or the wider market. To draw conclusions beyond the one process you studied, you need evidence for the wider scope. Narrow the wording where necessary.

Read it through a second time, looking at files and places where data could still be visible. Check chart captions, series labels, appendices, images, comments and tracked changes in the copy you have prepared. If you include a spreadsheet, look for extra sheets and comments. Compare the results you are allowed to include with the slides for your defence. If a monthly figure has been approved, check that no part of the presentation also shows a breakdown that was not agreed. If you use a label such as “Company A”, check that the description of the process and the rest of the material do not reveal the company's identity against the agreed terms.

Finally, prepare a short note for your supervisor. State what source you used, in what form, what was excluded and which limitation matters most for interpreting the results. In Ola's case, that would be the higher share of unknown outcomes in June. Also say what documents you can show and in what form. This makes it easier to discuss what the analysis can tell you and agree which sections need further work.

Sources

• Gaschi-Uciecha, A. (2025). Indicative assessment of the functioning of supply logistics in a selected distribution centre. Scientific Papers of Silesian University of Technology. Organization and Management Series, 237, 119-127. DOI: 10.29119/1641-3466.2025.237.7.
• Główny Urząd Statystyczny. (2 June 2026). Przewozy ładunków i pasażerów w 2025 r. Statistical release and tables.
• Makowska, M., and Boguszewski, R. (2013). Analiza danych zastanych – zagadnienia wstępne. In M. Makowska (ed.), Analiza danych zastanych. Przewodnik dla studentów (pp. 9-31). Wydawnictwo Naukowe Scholar.
• Politechnika Poznańska, Centrum Spraw Studenckich. (n.d.). Informacje dla dyplomanta – praca dyplomowa poufna.
• Rechciński, M., Balon, J., and Grodzińska-Jurczak, M. (2017). Dane zastane – ocena użyteczności do badania konfliktów społecznych wokół obszarów chronionych w trzech skalach przestrzennych. Prace Geograficzne, 149, 81-100. DOI: 10.4467/20833113PG.17.011.6927.
• Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 (EU GDPR), recital 26 and Article 4(5).
• Urząd Ochrony Danych Osobowych. (7 July 2025). Anonimizacja nie kończy się na zasłonięciu imienia i nazwiska.
• Ustawa z dnia 16 kwietnia 1993 r. o zwalczaniu nieuczciwej konkurencji, Article 11. Consolidated text: Dz.U. z 2026 r. poz. 85.
• Wyższa Szkoła Przedsiębiorczości i Administracji w Lublinie. (n.d.). Praca dyplomowa i obrona, section “Procedura objęcia pracy dyplomowej klauzulą poufności”.