“ESG reporting” on its own is not a research question — it is a subject heading, and a supervisor who sees it alone in a proposal has no way to tell what you would actually measure, where the data would come from, or what design fits. Every week spent defending a vague ESG topic instead of narrowing it is a week you are not writing Chapter 2, and a topic bounced back once at proposal stage typically costs a full revision cycle before ethics clearance can even begin.
Why does “ESG reporting” alone get sent back?
Because ESG reporting is a practice, not a problem. A defensible accounting dissertation needs to pick one of at least three distinct research jobs: measuring the quality or completeness of ESG disclosure itself (a content-analysis job), testing whether ESG disclosure or performance relates to a financial outcome like profitability or cost of capital (a correlational job), or explaining why some companies disclose more than others (a determinants job, often using company size, industry or ownership structure as predictors). These are three different studies with three quite different designs — naming which one you are doing, in the title itself, is the single biggest step toward a proposal your supervisor can approve quickly.
How do I write a research question that isn’t just “is ESG important”?
Take whichever of the three jobs above fits your interest, and add a sample, a time period and a specific metric. “Does ESG reporting matter?” is not yet a question; “does the extent of ESG disclosure in JSE-listed mining companies’ integrated reports predict their return on assets over a five-year period?” is. The second version names the sector, the disclosure vehicle (the integrated report), the outcome variable, and the time frame — everything an examiner needs to see the study is actually researchable within a normal dissertation timeline.
Which financial performance measure should the “outcome” side use?
If your research question tests ESG against financial performance, name the specific measure and be consistent about it throughout the dissertation rather than switching between measures informally. Return on assets and return on equity are the most commonly used accounting-based measures in this literature, since they can be calculated directly from a company’s own financial statements without needing a market-data subscription. Tobin’s Q or share-price-based measures are common in the finance literature but require market capitalisation data, which adds a data-access step your accounting-focused study may not need. Pick one primary measure, justify why it fits an accounting framing specifically, and treat any additional measure as a secondary robustness check rather than reporting several measures without a stated primary one.
Where does the data actually come from?
JSE-listed companies publish their own integrated annual reports publicly, and these — not a specialist commercial database — are the primary, freely accessible source for a South African ESG accounting dissertation. Corporate governance disclosure in South Africa is framed by the King Code on an “apply and explain” basis, meaning companies state how they apply each governance principle or explain why they do not: the King IV Report on Corporate Governance for South Africa (2016) was succeeded by the King V Code on Corporate Governance for South Africa, published by the Institute of Directors in Southern Africa (IoDSA) on 31 October 2025 and effective for financial years beginning on or after 1 January 2026. For a multi-year sample, record which code applied to each reporting year. The apply-and-explain logic is a useful frame for a determinants-of-disclosure study, since it itself creates variation in how much companies choose to say. Building your own disclosure index by reading each sampled company’s integrated report directly, rather than relying on a commercial ESG score you have not verified the methodology of, is the more transparent and more commonly accepted approach at honours and master’s level; commercial ESG ratings from providers used internationally are one option for a larger-sample quantitative study, but access and licensing cost are real constraints — confirm availability through your university library before designing a study around one.
A worked illustrative example
This example is entirely illustrative and fictional. Working title: “ESG Disclosure and Financial Performance Among JSE-Listed Retail Companies: A Content Analysis.” Research questions: RQ1 — What is the extent of ESG disclosure among a sample of JSE-listed retail companies, measured against a constructed disclosure index? RQ2 — Is there a relationship between the extent of ESG disclosure and return on assets over the study period? Population and sample: all JSE-listed companies in the retail sector with integrated reports available for the study period, an illustrative sample of 18 companies over five reporting years. Instrument: a disclosure index built from established ESG disclosure categories (environmental impact and targets, labour and safety practices, board diversity and governance structure, community and stakeholder engagement), each item scored present/absent or on a simple ordinal scale, applied consistently by the researcher across every sampled report — with a second coder checking a subset for consistency, exactly as a content-analysis study in any field would. Analysis: descriptive statistics for the disclosure index, then a correlation or regression test of disclosure score against return on assets, controlling for company size.
What method and analysis fit this kind of study?
A quantitative content-analysis design — building a disclosure index and scoring each sampled report against it — is the standard approach for the disclosure-quality and determinants jobs described above, and produces a dataset that ordinary statistical tests can then analyse. A qualitative design (thematic analysis of how companies frame their ESG narrative, or interviews with preparers about why disclosure choices are made) fits better if your interest is in meaning and motivation rather than measurable extent. Mixing both — an index score plus a smaller set of interviews explaining the pattern found — is a common and defensible master’s-level design where time allows for both halves.
What common mistakes weaken an ESG accounting proposal?
Three recur most often. First, mixing the three research jobs (disclosure extent, financial-performance relationship, determinants) into one vague study that tries to do all three without a clear primary question — pick one as primary and treat the others as secondary at most. Second, building a disclosure index by copying categories from an unrelated international framework without checking whether they fit South African reporting practice and the King Code’s apply-and-explain structure (King IV, or King V for financial years from 2026) specifically. Third, asserting a specific ESG statistic or trend (“ESG disclosure has increased by X% since Y”) without a named, opened source for that number — the same fabricated-statistic risk that applies to any field-specific claim, not unique to ESG.
What should I watch out for with ESG ratings and secondary data?
Commercial ESG rating providers do not always agree with each other on the same company, because they weight categories differently and are not fully transparent about their own methodology — a well-known limitation in the international ESG research literature, not a flaw specific to any one provider. If your design uses a commercial ESG score rather than your own disclosure index, name the specific provider, state the exact score date and methodology version you used, and acknowledge this rating-divergence limitation explicitly in your discussion chapter rather than presenting the score as an objective, uncontested measure.
How does this connect to other accounting content on this site?
If ESG framing doesn’t end up fitting your interests, the site’s 40 accounting and CA(SA) research report topics covers a wider spread of accounting sub-fields with their own research questions. For the statistical side of a quantitative ESG-and-performance design, the site’s guide to choosing an analysis method for a South African accounting research report covers the regression and correlation options in more depth, and the general guide to choosing the right statistical test applies if your design needs a different comparison. Where your ESG study also touches on AI-assisted analysis of large volumes of report text, the site’s guide to using AI safely in an accounting research report covers what is and is not an acceptable use of AI tools in that process — a separate question from the ESG topic itself. For the underlying company data beyond ESG specifically, the site’s guide to data sources for an accounting dissertation covers JSE, SARS and SAICA sources more broadly.
What if my supervisor still says ESG is “too trendy” or “already done”?
Bring the narrowed version — sector, sample, disclosure job, specific outcome variable — rather than defending ESG’s general importance. A supervisor rejecting “ESG reporting” as a heading is usually not rejecting the underlying interest; they are asking for the same narrowing any topic needs before it becomes a research question. If a very similar study has already been done on a different sector or market, naming that study and explaining precisely how your sector, sample or time period differs turns “already done” into evidence that your design is feasible and well-precedented, not redundant. A study that engages honestly with prior work, rather than ignoring it, is almost always judged more favourably by an examiner than one that claims, implausibly, to be entirely first.
Turning “ESG reporting” into a properly scoped research question, disclosure index and analysis plan — the exact structuring work examiners look for first — is where Tesify helps, while your sample, your index and every word of the write-up remain entirely yours.
Frequently asked questions
Do I need a commercial ESG database subscription to do this kind of study?
No — building your own disclosure index from publicly available integrated reports is a standard, accepted approach at honours and master’s level, and avoids the cost and methodology-transparency concerns of a commercial rating.
How many companies do I need in my sample?
This depends on your planned analysis — a purely descriptive content analysis can work with a smaller, deliberately selected sample, while a regression or correlation test needs enough companies for the statistical test to have adequate power; discuss the specific number with your supervisor once your analysis plan is set.
Is ESG reporting research considered accounting, or does it belong in management or finance?
It can sit in any of the three, depending on the specific angle — an accounting programme typically frames it around disclosure and reporting quality, a finance programme around the performance or cost-of-capital relationship, and a management programme around governance and stakeholder strategy. Confirm with your own department which framing fits its expectations.
Can I study a single company’s ESG reporting in depth instead of a cross-sectional sample?
Yes — a single-company or small-sample case study tracking ESG disclosure changes over several years is a valid alternative design, particularly useful where you want to trace how disclosure evolved around a specific event, such as a change in leadership or a sustainability-linked incident.
Do I need ethics clearance for a study based entirely on public company reports?
Studies using only publicly available secondary documents typically face a lighter ethics review than studies involving human participants, but most South African faculties still require some form of ethics application or exemption confirmation — check your department’s process rather than assuming a desk-based study is automatically exempt.
Should I compare South African ESG disclosure to international benchmarks?
You can, but be specific about which international study or benchmark you are comparing against and confirm the methodologies are actually comparable (same disclosure categories, similar sample composition) before drawing a direct comparison — a mismatched comparison is a common weak point examiners flag.