Explanatory Sequential Design Example in Nursing: Caring Behaviours

This explanatory sequential design example is a published study by Ferede and colleagues (2024) in BMC Nursing. The team first surveyed medical-surgical nurses in three Ethiopian referral hospitals about their perceptions of caring behaviours and the factors around them, and then interviewed a smaller group of the same nurses to explain what the numbers showed.

The page follows the study phase by phase: why the design was chosen, how the survey was built and analysed, how its results were used to plan the interviews, what the interviews added, and how the two sets of findings were brought together. It also includes a joint display built from the published results, a short critical appraisal and guidance on using the design in a nursing dissertation.

This explanatory sequential design at a glance

Aim

To examine factors associated with medical-surgical nurses’ perceptions of caring behaviours and to gain a more comprehensive understanding of them

Design

Sequential explanatory mixed methods (QUAN → qual)

Setting

Three referral hospitals in Amhara Regional State, Ethiopia: University of Gondar Comprehensive Specialized Hospital, Debre Markos and Debre Birhan Referral Hospitals

Data collection

January to March 2022

Phase 1

Census survey of nurses with a BSc or higher in internal medicine and surgical departments: 148 of 156 returned (94.9%)

Phase 2

Semi-structured interviews with 12 survey respondents chosen by maximum variation sampling

Ethics

University of Gondar Institutional Review Board, with permission from each hospital and written consent

Why an explanatory sequential design was chosen

The study had two linked needs, which is why an explanatory sequential design suited it. First, it needed to know which work-related and personal factors were associated with nurses’ perceptions of caring across a whole population of medical-surgical nurses, which calls for a survey and statistical analysis. Second, it needed to understand why those factors mattered in daily practice in Ethiopian hospitals, which numbers alone cannot show. Putting the survey first and the interviews second lets the quantitative results decide what the qualitative phase needs to explain.

• The quantitative strand comes first and carries the main weight. The survey identifies the associations; the interviews explain them.
• The phases are connected. Interviewees are drawn from survey respondents, and the interview analysis is guided by the survey findings.
• Integration happens at interpretation. Interview themes are used to explain and extend the statistical results.

How the explanatory sequential design connected the two phases

Phase 1 of the explanatory sequential design: the survey

Participants

Rather than drawing a sample, the researchers invited every nurse with a BSc degree or higher working in the internal medicine and surgical departments of the three hospitals. This census approach suits a defined, reachable population and removes sampling error within it. A total of 148 of 156 questionnaires were returned, a response rate of 94.9%.

Instruments

The questionnaire package was self-administered and was pretested with 15 nurses, about 10% of the sample, before the main data collection. Using a caring instrument that had already been translated into Amharic avoided the extra work and risk of a new translation.

Analysis and results

Multiple linear regression was used to identify factors associated with caring behaviour scores, and the authors report checking linearity, normality, homoscedasticity, multicollinearity and outliers. A two-sample t-test compared nurses who did and did not perceive resources to be available.

Nurses who felt that resources were available scored higher than those who did not (mean 4.93 compared with 4.69, p = 0.03). Together, these results told the researchers that staffing, stress, satisfaction and resources were linked to how nurses perceived caring, but not how or why. That gap is what the second phase was designed to fill.

Phase 2 of the explanatory sequential design: the interviews

Fifteen survey respondents, five from each hospital, were invited using maximum variation purposive sampling based on sex, education and experience. Three could not attend because of workload, so 12 nurses were interviewed. Choosing interviewees from the survey sample is the connecting step that defines an explanatory sequential design: the people who explain the numbers are the people who produced them.

1. Pilot interviewOne pilot interview tested the interview guide before data collection.
2. Face-to-face interviews in AmharicSemi-structured interviews were recorded with a digital voice recorder in the participants’ own language.
3. Translation and back-translationTranscripts were translated into English and blindly back-translated into Amharic to check accuracy.
4. Stopping pointData collection stopped after the twelfth interview, judged sufficient using the concept of information power.
5. Deductive codebook analysisThematic analysis following Braun and Clarke used a codebook guided by the quantitative results, so that themes could be matched to survey findings.
6. Checking the themesTwo researchers developed the themes together and senior researchers refined them; direct quotations support each theme.

Themes

Integration in the explanatory sequential design

The article integrates the two phases in its discussion, describing a connecting approach in which the qualitative findings explain and extend the quantitative results. The joint display below was prepared by our team from the published findings to show how this works in a format you can reuse in your own dissertation.

Read across each row and the value of the explanatory sequential design becomes clear. The survey shows that a factor matters; the interviews show how it matters in the wards; the final column is a conclusion neither phase could reach alone.

Critical appraisal of this explanatory sequential design

Reading a mixed-methods study critically is good practice for your own literature review. The authors identify the mixed-methods design, the inclusion of distant hospitals and the inclusion of all medical-surgical staff as strengths, and the relatively small quantitative sample and the single-item measures as limitations. Three further points are worth noticing.

Recommendations made by the authors

The authors recommended that health institutions adopt policies that allow nurses to assess their caring perceptions and working environment, that policymakers use systems-level approaches to support nurses and create better working environments, and that staffing and resource availability be addressed directly. Each recommendation can be traced back to a row of the joint display, which is a useful check on any mixed-methods conclusion.

Using an explanatory sequential design in your dissertation

An explanatory sequential design works well for a dissertation when a survey is feasible and you expect the results to raise “why” questions. Plan the connecting step from the start: decide how you will choose interviewees from your respondents and how the survey results will shape the interview guide.

Check feasibility before committing. A sequential design takes longer than a single-method study because the second phase cannot start until the first is analysed. It also needs two sets of skills: statistical analysis for the survey and qualitative analysis for the interviews. If your timetable is tight, keep the survey short, limit the regression to a few predictors you can justify from the literature, and plan a modest number of interviews chosen to explain the most important or surprising survey results. The three topics below show how the same two-phase logic can be applied to other nursing questions.

Sampling decisions in an explanatory sequential design

Sampling in an explanatory sequential design happens twice, and the two decisions have different purposes. In phase 1 the goal is coverage: the census of all eligible medical-surgical nurses with a BSc or higher made the survey results representative of that population in the three hospitals. In phase 2 the goal is explanation: maximum variation sampling by sex, education and experience made sure that the interviews heard from nurses whose survey answers were likely to differ.

Inviting five nurses from each hospital also kept all three settings visible in the qualitative data. When three invited nurses could not take part because of workload, the researchers stopped at twelve, using information power rather than a fixed number to judge sufficiency. In a dissertation, explain both sampling decisions separately and say how the second depended on the first.

Questions an examiner might ask, and how this study answers them

Working through likely examiner questions is a good way to check whether an explanatory sequential design example really holds together.

Why mixed methods at all?

The survey showed which factors were associated with caring; interviews were needed to explain why they mattered in practice.

Why this order?

The survey results determined what needed explaining, so the quantitative phase had to come first.

How are the phases connected?

Interviewees were drawn from survey respondents, and a codebook based on the survey results guided the interview analysis.

Were the survey measures reliable?

Cronbach’s alpha was 0.89 for the caring inventory, 0.82 for stress and 0.77 for the nurse-doctor relationship scale.

How were translation risks handled?

Interview transcripts were translated into English and blindly back-translated into Amharic.

Where does integration happen?

At interpretation, where themes are used to explain and extend the statistical findings.

A realistic timeline for an explanatory sequential design

Ethics applications for sequential designs should describe both phases, including how interviewees will be contacted from the survey sample. If your survey is anonymous, ask respondents to opt in to an interview by giving contact details separately.

Planning checklist for an explanatory sequential design

1. Write two linked questionsOne quantitative (“What factors are associated with…?”) and one qualitative (“How do nurses explain…?”), plus a mixed-methods question about how the second explains the first.
2. Choose tested instrumentsUse scales with published reliability, ideally already validated in your participants’ language.
3. Decide the connecting rule earlyState how interviewees will be chosen from respondents, for example by extreme scores or maximum variation.
4. Leave room in the interview guideDraft core questions in advance, then add questions that target the survey results once they are known.
5. Plan integrationSketch the joint display before analysis so you know what each strand must provide.

Writing the integration section

Integration is where many dissertations using an explanatory sequential design lose marks, usually because the two strands are reported one after the other with no attempt to combine them. A clear integration section has three parts. First, restate the key survey results in one or two sentences each. Second, for each result, describe what the interviews showed about it, using one short quotation where it helps. Third, state the combined conclusion, labelling it as confirmation, expansion or discordance: confirmation when both strands agree, expansion when the interviews add something new, and discordance when they disagree.

In this study, most rows of the joint display would be labelled as expansion, because the interviews explain the mechanisms behind the statistical associations. A row showing discordance is not a failure; it is a finding that deserves discussion, and examiners often look for evidence that you noticed it.

How to describe this explanatory sequential design in your literature review

Related examples and reviews

Each strand of this study links to other examples in our library. The survey strand works like the plan in our correlational research question example, and the interview strand can be compared with the in-depth approach in the phenomenology example. For the evidence on the topics behind the themes, see our literature reviews on burnout and staff wellbeing and therapeutic communication skills.

Common mistakes this explanatory sequential design avoids

• Two separate studies. The phases are connected through participant selection and the analysis codebook.
• Interviewing people outside the survey. All interviewees came from the survey respondents.
• Unclear notation. The order and weight of the strands are stated as QUAN → qual.
• Ignoring language. Translation and back-translation steps are reported.
• Integration left to the reader. The discussion explains how themes account for the statistical findings.

Use this explanatory sequential design example as a template: two linked phases, a clear connecting step and an integration table that brings the strands together.

References

1. Braun, V. and Clarke, V. (2006) Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), pp. 77–101. https://doi.org/10.1191/1478088706qp063oa
2. Ferede, A.J., Gezie, L.D., Geda, B. et al. (2024) Nurses’ perceptions of caring behaviors at referral hospitals in Ethiopia: a mixed-methods approach. BMC Nursing, 23, 767. https://doi.org/10.1186/s12912-024-02431-9