About this example
This page is a doctoral research proposal in business and management, written at the level expected for a PhD or Doctor of Business Administration (DBA) application or a first-year progression review. It proposes a quantitative survey study of how communication technology and management practices relate to the effectiveness of collaboration in remote and hybrid teams.
It is an illustrative teaching example, not a real student's work, and any organisation or participant details are anonymised or invented for teaching. The grey "What to notice" notes explain what each passage does and why.
Use it to see how a quantitative proposal links a problem to hypotheses, measures, sampling and analysis. Then build your own argument from your own reading. Do not submit any part of it as your own work; your university's academic integrity rules apply.
For the wider method, read the research proposal guide, selecting a PhD topic and choosing a dissertation methodology. More examples are in the sample work library.
The example: background and problem
Background
Working away from a shared office is no longer an exception for many knowledge workers. Arrangements that were introduced at scale during the COVID-19 pandemic have settled into a range of remote and hybrid patterns, and teams now coordinate much of their work through video meetings, chat platforms and shared documents. Research on virtual work has grown accordingly, and a recent review of global virtual work brings together findings on how distance, technology and team processes interact, while setting out questions that remain open (Froese et al., 2025).
The individual-level evidence is more settled than the team-level evidence. A meta-analysis of telecommuting studies found small beneficial associations with perceived autonomy, work-family conflict, job satisfaction, supervisor-rated performance and turnover intention, and no overall association with the quality of relationships with co-workers; however, high-intensity telecommuting (around two and a half days a week or more) was associated with poorer co-worker relationships (Gajendran & Harrison, 2007). That last finding points to collaboration, rather than individual productivity, as the place where remote work is most likely to create problems.
Research problem
Organisations continue to disagree about how much remote work to allow, and decisions are often made on managers' impressions of productivity rather than on evidence about how teams actually collaborate. Two sets of levers are within an organisation's control: the communication technologies it provides and the management practices it uses, such as goal setting, regular check-ins and how performance is monitored. Less is known about how these two sets of levers work together to support collaboration across different levels of remote intensity, or whether managers and employees judge collaboration in the same way. The problem this study addresses is the absence of an evidence-based account of which combinations of technology and management practice are associated with effective remote collaboration.
The example: aims, hypotheses and literature
Aim, objectives and hypotheses
The aim of the study is to explain how communication technology and management practices relate to the effectiveness of collaboration in remote and hybrid teams. The objectives are:
- To review research on remote work, virtual teams and communication technology, and derive a model of collaboration effectiveness.
- To measure employees' and managers' perceptions of communication technology fit, management practices and collaboration effectiveness.
- To test whether technology fit and management practices are associated with collaboration effectiveness, and whether remote intensity moderates these associations.
- To compare how managers and employees assess collaboration, and draw implications for organisational policy.
Three hypotheses follow from the literature. H1: perceived communication technology fit is positively associated with collaboration effectiveness. H2: supportive management practices are positively associated with collaboration effectiveness. H3: the association between remote intensity and co-worker relationship quality is negative, consistent with meta-analytic evidence (Gajendran & Harrison, 2007), and is weaker where management practices are more supportive.
Literature to be reviewed
The full proposal will review four areas: the history and outcomes of telework, using meta-analytic and review evidence rather than single case reports (Gajendran & Harrison, 2007); research on virtual and globally distributed teams (Froese et al., 2025); theories of communication media choice and fit; and management practices in remote settings, including goal setting, monitoring and trust. The review will end with a conceptual model showing the hypothesised paths, which becomes the structure of the questionnaire.
The example: method and ethics
Design and philosophy
The study adopts a post-positivist position and a cross-sectional survey design, which is suited to estimating relationships among measured variables in a defined population (Creswell & Creswell, 2023). It does not establish causation; the findings will be reported as associations, and the limitations of a single time point will be discussed.
Population, sampling and measures
The population is employees and their line managers in knowledge-work roles in organisations that operate remote or hybrid arrangements. Participating organisations will be approached through professional networks, and all eligible staff in participating units will be invited, so that managers and employees from the same teams can be compared. Constructs will be measured with established multi-item scales identified in the review, adapted where necessary and answered on a five- or seven-point agreement scale; remote intensity will be measured as the typical number of days a week worked away from the office. The questionnaire will be piloted with a small group for clarity and timing.
The minimum sample size will be set by a power analysis for the planned model (Faul et al., 2007), checked against simulation-based guidance for structural equation models, which shows that requirements depend on the number of indicators, factor loadings and effect sizes rather than on a single rule of thumb (Wolf et al., 2013).
Analysis and threats to validity
Data will be screened for missing values and outliers, and descriptive statistics reported. Treating multi-item Likert scale scores with parametric methods is defensible when scales are summed or averaged and samples are reasonable in size (Norman, 2010). The measurement and structural models will be estimated with structural equation modelling; if the model is complex relative to the sample, partial least squares SEM will be used and reported following current guidance (Hair et al., 2019), with discriminant validity checked using the heterotrait-monotrait ratio (Henseler et al., 2015). Manager and employee groups will be compared through multi-group analysis.
Because all variables come from the same questionnaire, common method bias is a risk; procedural remedies (separating predictor and outcome sections, protecting anonymity, varying response formats) will be used and statistical checks reported (Podsakoff et al., 2003). Online recruitment can produce self-selected samples that differ from the population (Bethlehem, 2010), so the sample profile will be compared with organisational data where available, and early and late respondents compared as a check on non-response bias (Armstrong & Overton, 1977).
Ethics
The study will be submitted for university ethics approval before data collection. Organisations will give written permission but will not see individual responses; participants will receive an information sheet, give consent online, may skip questions or withdraw before submission, and will not be identifiable in any report. Data will be stored securely in line with data protection law. A DBA candidate researching their own organisation should also reflect on how their role shapes access and interpretation (Farrell et al., 2018).
The example: timeline and expected contribution
Timeline
| Period (full-time) | Main activities | Output |
|---|---|---|
| Months 1 to 9 | Literature review; conceptual model; scale selection; ethics application | Review chapter; approved questionnaire |
| Months 10 to 15 | Pilot; organisational access; main survey | Cleaned data set |
| Months 16 to 24 | Measurement and structural models; multi-group analysis | Results chapter draft |
| Months 25 to 36 | Discussion, implications, writing and submission | Thesis; practitioner summary for participating organisations |
Expected contribution
Theoretically, the study will test whether technology and management practices act as complementary levers for remote collaboration, and whether supportive management weakens the negative link between remote intensity and co-worker relationships reported in earlier work (Gajendran & Harrison, 2007). Practically, it will give organisations evidence on which combinations of tools and practices are associated with better collaboration, and on where managers' and employees' views diverge. For DBA candidates, this practical contribution is often as important as the theoretical one.
How to adapt this to your own study
- Replace headline statistics with evidence from reviews or meta-analyses, and describe effect sizes honestly.
- State a problem as a decision being made without evidence, and name the variables you can measure.
- Write objectives your design can meet; a cross-sectional survey tests associations, not causes or futures.
- Derive each hypothesis from the literature and draw the conceptual model before writing questions.
- Choose established scales and justify your sample size with a power analysis for your actual model.
- Plan for common method bias, self-selection and non-response before you collect data.
- If you research your own organisation, explain how your role affects access, responses and interpretation.
Common mistakes in this chapter
- Opening with impressive statistics from news sites or blogs that cannot be traced to a primary source.
- Using company anecdotes as evidence for a general problem.
- Promising prediction or causal conclusions from a one-off survey.
- Measuring "productivity" by self-report without acknowledging the limitation.
- Saying "a Likert questionnaire will be used" without naming constructs, scales or sample size.
- Ignoring common method bias when all variables come from one survey.
- Listing a "plagiarism" section or other service content inside an academic proposal.
| What examiners look for | Where this example does it | Check in your own draft |
|---|---|---|
| Evidence-based background | Meta-analysis and review instead of unsourced figures | Can every number be traced to a source? |
| A focused problem | Decision made without evidence; two sets of levers | Does the problem name measurable variables? |
| Achievable objectives | Associations tested, not futures predicted | Can your design meet every objective? |
| Hypotheses grounded in theory | H1 to H3 linked to the review | Is each hypothesis traceable to a source? |
| A justified sample | Power analysis and SEM guidance | Have you calculated, not guessed, your sample size? |
| Planned analysis | SEM or PLS-SEM, HTMT, multi-group analysis | Does each hypothesis have a test? |
| Awareness of bias | Common method, self-selection and non-response checks | Which biases threaten your design? |
| Workplace ethics | Organisations do not see individual answers | How will you protect employees from their employer? |
Key takeaways
- Build the background on reviews and meta-analyses, and report effect sizes as they are (Gajendran & Harrison, 2007).
- Write objectives your design can actually meet; a survey tests associations, not causes.
- Justify the sample size for the model you will estimate, not a rule of thumb (Wolf et al., 2013).
- Plan procedural and statistical responses to common method bias before fieldwork (Podsakoff et al., 2003).
- DBA candidates researching their own organisation should make their insider role part of the method (Farrell et al., 2018).
