Analysing logistics complaints in a master's thesis. How do you compare case counts and costs?

You already have a company complaints register without information that identifies customers. Delays account for the most cases in your spreadsheet, while the largest amounts appear under damage or loss. Should delays or damage come first in your results chapter? That depends on whether the question concerns case counts, documented direct costs or the number of complaints relative to the number of shipments dispatched. Each question needs its own numerator and denominator.

Published: Updated: About 26 minutes to read
Cardboard parcels and a printed register on a dispatch-area worktop
Cardboard parcels and a printed register on a dispatch-area worktop.

You already have a company complaints register without information that identifies customers. Delays account for the most cases in your spreadsheet, while the largest amounts appear under damage or loss. Should delays or damage come first in your results chapter? That depends on whether the question concerns case counts, documented direct costs or the number of complaints relative to the number of shipments dispatched. Each question needs its own numerator and denominator.

This guide takes you from the register to three sets of results and shows you how to structure the paragraphs that explain them. The distribution company is fictional and all figures are illustrative. For your own data, start by defining what counts as one case, which shipments you can compare and what the amount column represents. Those decisions will help you explain exactly what you counted when your supervisor asks about one of your charts.

Establish which part of the process the register covers

A complaints register documents cases that reached a particular channel and were recorded. A company may separately track errors caught in the warehouse, requests for status updates and complaints filed after dispatch. If you received only the last register, frame your research question around recorded complaints. To estimate how often shipments were actually late, you would need both the planned or agreed date and the actual date for the event you are analysing (dispatch or delivery), including for shipments that received no complaints.

Turn the aim into three questions you can answer with the data

An example research aim is: “The analysis compares categories of recorded complaints about shipments in a defined distribution process by case count and documented direct cost. It also compares the number of complaints per 1,000 shipments across two dispatch periods.” In the methods chapter, break this aim into questions that lead to separate results.

  1. How frequent is each category in the register? Count the cases in each category and calculate their share of all eligible cases. The answer describes the composition of the register.
  2. Which categories account for the greatest total recorded direct cost? Add the available, verified cost entries within each category. Report the amounts alongside the number of cases with complete, partial or unavailable cost records.
  3. How many cases are there per 1,000 shipments in each period? Link complaints to the corresponding shipments, establish a comparable follow-up window and divide the number of complaints by the number of shipments observed.

A heading such as “Analysis of complaint causes” is suitable when you have evidence of verified causes. If the export contains the customer's account, a category and a decision, call the section “Complaint categories and documented costs”. The category supports a statement such as “The complaint recorded damage.” A statement that the carrier damaged the goods requires evidence showing when and how the damage occurred. Even a note that a complaint was upheld records a decision in the company's system. Keep the decision, the category and any finding about the cause in separate fields.

Choose the unit of analysis and use it throughout the chapter

The unit of analysis in this example is a complaint case linked to one shipment. A complaint may have several updates, attachments and cost lines under the same identifier. If the register contains a row for every status change, a simple row count will give extra weight to cases that take longer to resolve. Before building a summary, count unique identifiers and inspect some repeated entries to see what they represent. Identical records in overlapping exports and successive handling events for one case call for different decisions in the spreadsheet.

A shipment may have several distinct cases, and one combined case may cover more than one shipment. In the first situation, complaints per 1,000 shipments still measures the number of cases relative to shipping volume. To calculate the percentage of shipments with a complaint, count the unique shipments linked to at least one case. Assign combined cases using the underlying records. If you cannot determine the shipments involved, report how often this occurs and how you treat those records in each analysis. The case can still be a suitable unit for a breakdown of categories and costs, while the shipment-based numerator remains incomplete.

Decide whether to include cases that are rejected, withdrawn or still open. If you are studying the workload associated with complaints, different statuses can be relevant because each eligible case reached the team. If your purpose is to compare upheld complaints, restrict the numerators to those cases and use the same definition in the result heading. When studying the mix of complaints received, an unavailable amount or a pending decision at the cut-off date leaves the case eligible for the count.

Prepare a dataset that lets you reproduce the calculations

Company data will often give you three usable sets of fields. The first describes cases, the second documents cost entries and the third identifies shipments. If you have one spreadsheet, grouping its rows can give you these three views, provided the identifiers and variables support the split. Keep the received file intact in the location agreed with the company and use a separate copy within the authorised scope. Your thesis needs the fields that explain the analysis and the aggregates you may disclose.

Link records at the level where they were created

  • Cases. A working case identifier, shipment key, first report date, original category, analysis category, status at the cut-off date and, where relevant, closure date. The financial status can be separate from the case-handling status.
  • Costs. A unique entry key or another verified way of identifying a single entry, case identifier, cost type, posting date and amount on a consistent basis. Keep a correction linked to the entry it adjusts. Use a documentation status to distinguish partial from final amounts and verified zero costs from blanks.
  • Shipments. A working shipment key, dispatch date under your chosen definition, process scope and, if needed and available, a comparable service segment. Use this dataset to count shipments that never generated a complaint as well.

Before joining cases and costs, aggregate cost lines to one amount per case and assess documentation completeness separately. If a case has three updates and two cost lines, joining every update to both lines produces six rows. Each amount now appears three times. Compare the number of cases before and after the join, and reconcile costs in the case view with the sum of the eligible lines. Correct date and decimal separator problems while preserving the meaning of the fields. If the file stores amounts as text, test the conversion on several actual formats before recalculating the entire spreadsheet.

If some cases lack a shipment key, give the match count and explain the reasons your authorised data support. Suppose the company has complete shipping records and you can match only 80 of 100 complaints. Dividing 80 by the full shipment volume describes only matched cases. If missing keys are concentrated in one period, the ranking of periods could change even in a comparison restricted to matched cases. With just the complaints register, you can answer questions about category composition and available costs. A rate per 1,000 shipments requires volume data that correspond to its numerator.

Define the categories before counting

Create a short coding guide with the category name, inclusion rule, a borderline example and a decision for cases involving more than one issue. The fictional example uses five main categories: delay, damage, incomplete delivery, loss and other complaints with a known issue. Delay covers timing. Damage concerns the condition of the shipment or contents. An incomplete delivery means part of the contents arrived. Loss means that the entire shipment did not arrive and the register applies that code. The “other” category covers remaining identifiable problems, such as errors in delivery paperwork. A real record lacking the information needed for classification can stay “undetermined”, keeping missing codes visible. All 100 cases in the example have a category.

Test the guide on entries from different weeks and statuses before assigning every code. Think of this report: “The shipment arrived a day late and part of the contents was damaged.” If the shares in your results are to total 100%, choose one main code using a stated rule, such as the issue marked as primary on the complaint form. Keep a secondary code in a separate field if useful. When the order in which the customer mentions the issues is the only evidence, explain how you deal with cases where the primary issue is unclear. If a supplementary analysis counts all issues mentioned, one case can contribute to several categories. Each share then represents the percentage of cases in which an issue appeared, and the shares may add up to more than 100%.

Preserve the original code and record the mapping. “Late delivery notice” and “late delivery” can share a category when the definition you use covers both. If code names changed when a new system was introduced, check that their meanings are comparable on both sides of the change. If some codes are incompatible, show the periods separately or use a common, broader category. In the methods chapter, state which version of the coding guide you used and how many entries required an individual decision. Explicit mapping rules make it easier to tell whether the eventual prominence of a category reflects the data.

Keep a short record of your checks

Once the rules are set, trace the number of source rows through to the number of unique cases. Record how many rows were updates, how many cases were outside the process or period, and how many cases lacked a shipment key. Check the relevant count in each resulting view. The breakdown by category uses all eligible cases, a cost-share chart uses the total of documented entries, and a period rate needs the full shipment volume for the same scope. A dated record of the rules and counts lets you repeat the analysis after an export is corrected.

Before presenting results, consider whether a combination of published details could reveal the company, a customer or a transaction, especially when an unusual case involved a large amount. The authorised scope may allow totals by category while excluding the date of an individual case, the route, a document number or a price that could identify the event. If you need to revisit how to describe the materials the company shared, see the guide to using company data in a master's thesis. Here, the scope of the source data sets the boundaries for the calculations.

Separate the case count, documented cost and rate per 1,000 shipments

The example covers shipments from the same distribution service in two consecutive months, A and B. Each shipment is assigned to a period by its dispatch date. The study looked for cases first reported within 30 days of dispatch, starting on the dispatch date and ending 29 days later. A complaint entered during B may therefore concern a shipment from A. The dispatch date controls the period assignment. The data were exported after that observation window had elapsed for every shipment. Costs and case statuses were taken at one common cut-off date, which can still leave cases from the two periods with different lengths of time for costs to be posted.

There were 2,000 shipments in A and 4,000 in B. Within the observation window, 45 and 55 cases respectively were linked to those shipments. Every case in this example concerns a different shipment, and the five primary codes cover the same 100 cases. The counts in parentheses are for A and B respectively.

  • Delays. 48 cases (20 and 28), or 48% of complaints. Documented direct costs were PLN 1,200.
  • Damage. 25 cases (12 and 13), or 25% of complaints. Documented direct costs were PLN 4,200.
  • Incomplete deliveries. 15 cases (7 and 8), or 15% of complaints. Documented direct costs were PLN 900.
  • Loss. 6 cases (3 and 3), or 6% of complaints. Documented direct costs were PLN 3,600.
  • Other. 6 cases (3 and 3), or 6% of complaints. Documented direct costs were PLN 100.

The category counts add up to 100 cases, shipment volumes to 6,000 and amounts to PLN 10,000. The cost figure includes all documented entries available at the cut-off date, including entries for partially reconciled cases. It gives you the breakdown of costs visible in the spreadsheet. The number of cases still lacking complete cost records limits any interpretation of the final costs.

How can you distinguish complete records, partial amounts, zero and missing data?

Agree on a cost definition with someone who knows the company's records. In the example, direct cost is the sum of costs that can be documented, were posted by the cut-off date and can be assigned to a specific case. The definition includes additional transport, return or replacement shipping costs, and the documented value of replaced goods, calculated on a consistent accounting basis. All entries are in Polish zloty. The amount requested by the customer, the selling price and an estimate entered as a provision belong in separate fields. If you want to analyse the cost ultimately borne by the company after carrier reimbursements, define this as a separate measure. Explain when reimbursements enter it and reconcile it with the amount before recovery. The main measure in the example excludes recoveries.

Before adding entries, check whether two records describe the same cost. A consolidated service invoice and a line attributed to a particular shipment may cover the same trip. If a cost covers several cases and the company shares only the combined total, that amount alone cannot establish the cost of each case. You can report the documented total outside the category breakdown and say what the breakdown covers. An estimated allocation requires a stated basis and separate presentation. Staff time, lost sales and reputational damage need a separate method if your study includes them. Keep estimates labelled separately from posted direct costs.

Record the financial position using distinct groups:

  • Complete record with a cost. The documentation has been confirmed as complete for the chosen cost definition. The total includes the documented entries and their corrections.
  • Complete record with a confirmed zero. A check established that there were no cost entries within the scope of the analysis. The case has a cost of PLN 0 and counts in the denominator for the average of complete cases.
  • Partial record. At least one eligible amount is documented and the complete cost is still unknown. Record the amount available at the cut-off date and the documentation status.
  • No documented amount. The ledger is blank and does not verify a zero. Include the case in the count of complaints and report its cost as missing.

Case-handling status and cost status are separate. An open case may already include an invoice for a replacement shipment. A closed case may still lack a linked invoice or a verified zero. If the system automatically fills blanks with zeros, check how the amounts were verified on a few records. Treating missing amounts as zero will lower the average and inflate the completeness rate even if the sum of recorded amounts stays the same.

In the example, 90 cases have complete records, including 12 confirmed zeros. A further 4 have documented partial amounts and 6 have no recorded amount. Eight cases are open in the handling system: four with partial costs and four with no amount. The other two without amounts are closed in the handling system and still lack complete cost documentation. Partial amounts total PLN 1,200, and complete records total PLN 8,800. The category breakdown shows where to exercise most care when interpreting costs.

  • Delays. 45 complete records totalling PLN 1,100, including 8 confirmed zeros. One partial entry of PLN 100 and two cases with no amount.
  • Damage. 21 complete records totalling PLN 3,700, including one confirmed zero. Two partial amounts totalling PLN 500 and two cases with no amount.
  • Incomplete deliveries. 14 complete records totalling PLN 900, including two confirmed zeros. No partial amounts and one case with no amount.
  • Loss. Four complete records totalling PLN 3,000, including one confirmed zero. One partial amount of PLN 600 and one case with no amount.
  • Other. Six complete records totalling PLN 100. This category has no confirmed zeros, partial amounts or missing amounts.

Cost documentation is therefore complete for 90% of all cases. It is complete for 45 of 48 delay cases (93.75%) and four of six loss cases (66.67%, rounded). That difference matters to the comparison. You can report average cost among cases with complete documentation in each category. Give the count and scope in the label. For loss, it is PLN 3,000 / 4 = PLN 750 per case with complete records. Dividing PLN 3,600 by six gives PLN 600 of documented cost to date per recorded case in the category. That measure includes a partial amount in the numerator and a case with no amount in the denominator. Show both measures if they answer separate questions, and label each clearly.

How can you show the change in ranking by case count and cost?

To find a category's share of all cases, divide its count by 100 and multiply by 100%. To find its share of documented cost, divide its total cost by PLN 10,000 and multiply by 100%. Damage accounts for 25% of cases and 42% of costs, loss for 6% of cases and 36% of costs, and delays for 48% of cases and 12% of costs. Incomplete deliveries account for 9% of costs and other complaints for 1%. The cost denominator covers only documented amounts at the cut-off date. State under the chart which partial records were included.

Choose two bar charts. In the first, order categories by case count and give their counts and shares, for example “48 cases, 48%”. In the second, order them by documented amounts and indicate that the unit is PLN. Use consistent names and colours if readers are following the position of each category. You can also keep the first chart's category order in the second when the aim is to make direct comparisons across the same positions. Descending order by amount is needed if you add a cumulative cost-share line. With only a few categories, two well-labelled charts can answer the questions on their own.

If you use a Pareto chart, add the shares after sorting by the chosen measure. By frequency, delays and damage account for 73% together. The three largest categories by count, including incomplete deliveries, account for 88%. By cost, damage and loss account for 78% of the documented total, or 90% when delays are added. Loss and other complaints have equal counts. You can display them in the order used in the coding guide and mention the tie. If you use an 80% threshold, explain it as a presentation choice. In the example the first two cost categories reach 78% and the first three reach 90%. The threshold does not by itself determine which measures the company should take.

A Polish study by Miller and Kaniowski (2007, pp. 74-78) illustrates this separation. The authors presented separate rankings by defect frequency and the costs associated with products complained about in the pharmaceutical company they studied. Their approach supports the design of the comparison. The numbers in this guide come from a separate, fictional logistics example. When using the literature, distinguish the source of the data from the way figures were calculated. Bogacz and Migza (2015, pp. 5591-5593) describe limitations in the way a manufacturing and distribution group recorded returned goods and propose linking case classification with sales, logistics and production data. Rudzewicz (2015, pp. 51-52) reports a breakdown of complaints by reason and separate measures of complaint handling at the company studied. Such studies can justify your choice of fields and measures where they address a matching problem in your own project.

How do you compare periods with different shipment volumes?

Compute the rate per 1,000 shipments by dividing the number of eligible complaints linked to shipments from a period by the number of all eligible shipments from that period, then multiplying by 1,000.

  • Period A. 45 / 2,000 × 1,000 = 22.50 cases per 1,000 shipments.
  • Period B. 55 / 4,000 × 1,000 = 13.75 cases per 1,000 shipments.
  • Both periods combined. 100 / 6,000 × 1,000 = 16.67 cases per 1,000 shipments, rounded.

Period B had 10 more cases than A. Shipment volume doubled and the rate fell by 8.75 cases per 1,000 shipments. The combined rate of 16.67 comes from adding the numerators and denominators across periods. The period with 4,000 shipments therefore contributes more to the combined result. A simple average would give the two periods equal weight. Use the decimal mark appropriate to your thesis language and apply a consistent number of decimal places where this helps the reader compare values.

You can calculate the same rate for one category when the same shipments were covered by its reporting process. For delays, the rate in A is 20 / 2,000 × 1,000 = 10.00 and the rate in B is 28 / 4,000 × 1,000 = 7.00. For loss, there were three cases in each period. With denominators of 2,000 and 4,000, the rates are 1.50 and 0.75 cases per 1,000 shipments respectively. Show those rates alongside the raw counts so readers can see that a visible difference in the rates rests on a small number of events.

Before comparing the periods, check that shipments and complaints were recorded under the same definitions. If A included only domestic shipments and B included international shipments as well, a difference could reflect a change in the mix. If the data allow, compare the same segment and give each volume. Check whether a new form became available or the rules for registering complaints by phone changed during one of the months. A rise or fall in recorded cases can accompany a change in reporting. Without this information the result describes the register within the stated scope.

If you have only the number dispatched in a month and the number of complaints filed in that month, check how many complaints relate to those shipments. Complaint intake by filing date helps describe team workload. A complaint rate for a shipment cohort links the numerator to the shipments in the denominator. Comparing two cohorts in a small study can be easier to justify than lining up several months taken from unrelated reports.

Check which results depend on individual cases

The loss category contains six cases, including one large case with a documented cost of PLN 2,000. It accounts for 20% of the total documented cost and more than half of the category's cost. First reconcile the entry with the source: verify the case assignment, units and uniqueness. If it is correct, retain it in the main results. Then you can test how much the cost breakdown depends on that case.

Subtracting PLN 2,000 from the two relevant totals for a sensitivity check leaves PLN 8,000 in documented costs, of which PLN 1,600 relates to loss. The loss share is now 20%. The damage share is 52.5%, with its numerator still at PLN 4,200. The main results were 36% and 42% respectively. The ranking of the two categories stays the same, while the lead for damage grows. Label this as a sensitivity analysis and state the amount subtracted. The main case counts and rates per 1,000 shipments still use all observations.

You can run a second check without removing valid cost entries. Limiting the totals to cases with complete records again puts damage first (PLN 3,700), loss second (PLN 3,000) and delays third (PLN 1,100). This shows the same order across the two definitions of available costs. Amounts still missing from the documentation may change the final ranking. In particular, documented costs for damage and loss differ by PLN 600, and two loss cases lack complete records. There is no basis for assigning those cases specific amounts. Report the cut-off date and the extent of the missing information.

For the smallest categories, give counts and put changes in context. When there were three cases in each period, one extra recorded case at the same volume would move the rate in A by 0.50 per 1,000 and the rate in B by 0.25 per 1,000. This tells readers how sensitive the comparison is to one event without making a claim about how often it will happen in future. If you use “significant” in the statistical sense, plan a test or confidence interval suited to the unit of analysis and any dependence between observations. For a descriptive account of the register, counts, differences and a clear scope are sufficient. Applying the result to future shipments requires separate justification.

If a small count paired with a large amount could identify a company, route or shipment, ask the person responsible for the material which aggregates can appear in the thesis. Hiding a single figure may leave it recoverable by subtracting the other categories from the total. Combining categories or reducing detail with the company's agreement can protect information in the material. In the methods chapter, state what was aggregated and what that changes about the way the results can be read.

Give methods, results and discussion distinct jobs

The methods chapter explains how the figures were produced. The results report the values, their ranking and any incompleteness. The discussion relates the findings to the literature, states limitations and justifies further investigation. Complaints, shipments and zloty amounts may appear in consecutive sentences. For each number, make it clear what it measures. The example paragraphs below still refer to the fictional data.

Example methods paragraph

The unit of analysis was a complaint case linked to one shipment dispatched through the selected distribution service. Two shipment cohorts were assigned to periods A and B by dispatch date. Cases were included if they were first reported from the date of dispatch to the 29th day after dispatch inclusive, regardless of their status at the cut-off date. Each case was assigned one primary category using the coding guide and the recorded code for the main issue. Direct costs were measured as the sum of eligible posted entries assigned to a case by a common cut-off date. Partial amounts, confirmed zeros and cases with no documented amount were recorded separately. The period rate was calculated by dividing the number of cases linked to a cohort by the number of shipments in that cohort and multiplying by 1,000.

In your thesis, add the actual dates or period labels you are allowed to disclose, the source and time of export, the number of unmatched cases and rules for identifying repeats. If a follow-up window is defined using a delivery date and the register contains losses, explain which dates you use for shipments without confirmed delivery. A uniform window starting at dispatch resolves that problem in this example. Check any other definition against all categories and explain its consequences. Also specify whether every shipment had the same reporting coverage in the channel studied.

Example results paragraphs

Among 100 recorded cases, delays were the most frequent category (48 cases, 48%). Damage (PLN 4,200 of PLN 10,000, 42%) and loss (PLN 3,600 of PLN 10,000, 36%) accounted for the largest shares of documented direct costs. Complete cost documentation was available for 90 cases, including 12 with a confirmed zero. The documented total also included PLN 1,200 from four partial records. Cost amounts were unavailable for six cases.

Period A recorded 45 cases across 2,000 shipments, and period B recorded 55 across 4,000 shipments. The rates per 1,000 shipments were 22.50 and 13.75 respectively. From A to B the rate fell by 8.75 cases per 1,000 shipments while the case count rose by 10. Each period had three loss cases, equivalent to 1.50 and 0.75 cases per 1,000 shipments respectively. The rates cover complaints recorded in the same follow-up window after dispatch.

Start a chart description with the comparison it permits. For the first example paragraph, use one chart titled “Case counts by category” and another titled “Documented direct costs by category”. For the rate chart, explain in the caption that cases are grouped by shipment dispatch date and give the length of the observation window. State the main point in the text so it is understandable without reading individual bars.

Example discussion paragraph

Categories ranked by case count occupied different positions when ordered by documented direct costs. Delays accounted for the largest share of cases. The higher amounts for loss despite its lower case count show that the choice of category for further investigation depends on the measure. The share of cost attributed to loss was sensitive to one PLN 2,000 entry and to the completeness of records at the cut-off date. The lower rate in period B refers to reported cases for the specified cohort and within the defined observation window. Further investigation could compare similar shipment segments, examine consistency of reporting and follow up missing cost entries.

If you want to suggest an area for improvement, tie it to the result and identify the evidence needed. A large number of delay complaints may justify checking how delays vary with order acceptance time and planned delivery date. A high documented cost for damage could justify checking whether costs are concentrated in a particular type of goods or packaging method. Both are further lines of investigation. Evaluating the effect of a process change requires a design that can compare conditions before and after the change and account for other changes in the same period. Attributing responsibility calls for evidence about the specific event and a separate assessment.

Check that the comparison can be reproduced

Before submitting the chapter, pick one value from each result. Reproduce the 48% for delays from the unique case count. Reproduce the 42% cost share for damage using the entries posted by the cut-off date. Reproduce the 13.75 rate for period B using cases linked to its cohort and the full shipment volume of that cohort. If you cannot trace a result back to those fields, add the missing definition to the methods or correct the formula in the working copy.

  • Can you see when successive records belong to the same case and how a complaint involving several issues was handled?
  • Do the main categories total the unique case count, and do the eligible cost entries reconcile to the amount reported?
  • Does the documentation status distinguish complete records, confirmed zeros, partial amounts and missing data?
  • Do both periods count shipments and linked cases using the same rules and follow-up window?
  • Do charts state the denominator, unit and cut-off date wherever the result depends on them?
  • Are raw counts visible for rare cases, and can the reader see how an unusually large amount affects the analysis?
  • Does the discussion state the limits of the data and what would be needed to investigate further when the register shows only a category or a difference between periods?

For a meeting with your thesis supervisor, bring a short account of sources and decisions, the three main summaries and a list of missing fields. One question you might ask is “Should we present all documented amounts as the main cost result and add a separate version limited to cases with complete records?” With the summaries ready, it is easier to assess how the choice answers your research question. If you have only available costs and the breakdown of cases, present those results and explain what you would need for the shipment rate.

Questions about complaint analysis and preparing your material

How does thesis scope differ from study limitations?

Scope defines the deliberately chosen problem, evidence, population and period. Limitations explain how design features or difficulties restrict interpretation. An omitted topic is not automatically a flaw; the important question is whether the conclusions stay within the stated scope.

How can I connect a limitation to its effect on the conclusions?

Identify a specific feature of the procedure, the finding or claim it affects, and the consequence. Then explain what you actually did to address it and what remains uncertain. A generic list mentioning sample size and time pressure is not enough.

Can an already prepared research chapter be improved?

Yes. I can check the consistency of tables, charts, calculations, comments, and conclusions, and adjust the layout to the supervisor's comments. To verify the accuracy of the results, I will also need the source data and the research instrument.

How does preparing the research chapter differ from statistical analysis?

The research chapter combines a description of the actual study, presentation of results, interpretation and answers to the thesis research questions. Statistical analysis concerns the choice of methods and calculations on quantitative data. You can request both types of support together, but I agree any new tests and verification of calculations separately after reviewing the material.

Do I need to have collected data already to prepare the research part?

To describe the results, actual research material is required. Before collecting data, I can help plan the method, tool, and structure of the future analysis. I do not create results or descriptions of participants who did not take part in the study.

What should I send for a quote?

Provide your topic, subject area, type of thesis, stage of preparation and required deadline. Describe what needs to be prepared, checked or revised. Attach the current document, university guidelines and any comments from your supervisor.

For research support, include your research outline, survey or interview guide and the data you have collected. For editing that addresses text similarity, attach your JSA report if you have one. You don't need every document before asking about the scope of support.

Sources

  • Bogacz, P., Migza, M. (2015). Budowa wielokryterialnego systemu ewidencji i analizy reklamacji dla grupy przedsiębiorstw produkcyjno-dystrybucyjnych. Logistyka, 3/2015, 5590-5594.
  • Miller, P., Kaniowski, G. (2007). Wykorzystanie analizy Pareto-Lorenza w badaniu przyczyn reklamacji i działaniach zmierzających do doskonalenia jakości w przedsiębiorstwie. Studia i Prace Kolegium Zarządzania i Finansów, 79, 72-90.
  • Rudzewicz, A. (2015). Efekty zarządzania logistyczno-marketingowego w przedsiębiorstwie. Zeszyty Naukowe Uniwersytetu Szczecińskiego, 850, Problemy Zarządzania, Finansów i Marketingu, 37, 45-54.

Would you like help presenting complaint analysis in your logistics thesis?

Prepare your topic, research question, a description of the source registers and summaries you may share. Tell us whether you need help with calculations, choosing charts, or explaining the findings and their limits.

Analysis and presentation of research results for your thesis
Contact StudentSupport

Related articles

Bachelor's and master's thesis in resocialisation - research and conclusions

Use programme documents and interviews in a resocialisation thesis: choose a question, analyse the accounts and distinguish programme delivery from effectiveness. How can you plan a resocialisation study when you have a programme description and access to interviews with educators? First work out which question these sources can answer. Staff can discuss how sessions are organised and what makes support difficult to provide. To describe participants' experiences or what happens to them later, however, you need information from them or other appropriate sources. “The effectiveness of resocialisation” might sound like a useful starting point, but what exactly do you want to examine? The organisation of support, experiences after release or a change in particular behaviour? A few staff interviews can reveal organisational difficulties without showing how participants' lives developed afterwards. This guide uses preparation for seeking employment after release as an example. I will show what documents and interviews can establish, and which questions they leave unanswered. Before choosing a title such as “The effectiveness of resocialisation in prisons”, list a few things you can investigate. How are participants prepared to access support after release? What information do they receive? What makes referral to another organisation difficult? These questions concern support arrangements and interviewees' experiences. They do not describe every change in a person's life, but they give you a specific project to develop. Suppose you can read the description of an employment-preparation programme, obtain aggregate delivery records and interview the staff running sessions. You could ask: “How was employment preparation organised, and what difficulties do staff identify?” Do not also promise to establish whether participants found work after release if you cannot obtain follow-up information. Session records cannot replace it. Naming a facility does not yet explain what you are studying. It identifies where you may collect information. Your question might concern information provision, cooperation with other organisations or how participants understand the support offered. Specify the period and whose experiences interest you. Otherwise, a staff member may discuss a different programme edition from the one covered by your documents. List the documents and people you may be able...

Read article

Purposive sampling for sociology interviews - how to report a change in recruitment

You have sent out interview invitations, but most of the people who respond completed the programme. Your sociology thesis at a university in Poland also needs to cover people who left early. You have sent out interview invitations, but most of the people who respond completed the programme. Your sociology thesis at a university in Poland also needs to cover people who left early. Before booking more interviews, identify the gap in your sample and work out how to report it in your methodology chapter. Separate eligibility criteria from recruitment channels. The criteria tell you whether someone's experience is relevant to your research question. A channel determines how the invitation reaches them. If an important perspective is missing, first check whether your current channel is likely to reach the people who could offer it. In this fictional master's project, a student asks how people who completed a first-year peer mentoring programme and people who left early describe their decision to stay or leave. She studies one programme over one semester. She is interested in the experiences of participants who had already met with a mentor. At the planning stage, a short worksheet is enough. Record the criterion, how you will check it, whose perspective is missing and which channel you could try next. Initial screening in this example only needs self-confirmation of being at least 18, the programme round, confirmation of at least one meeting and self-reported completion status. The reason for leaving comes up in the interview, so it is not a condition for being invited. Palinkas and colleagues describe criterion sampling and strategies that deliberately capture relevant differences between cases. An announcement in the programme's active channel led to five interviews with people who completed the mentoring and one with someone who left. That interview describes a timetable change. The material is too narrow to address the different circumstances of leaving that the project asks about. Reposting the same invitation in the same place may keep favouring people still connected to the programme. Decision log after six interviews. With the channel administrator’s permission, a second invitation went out...

Read article