ILLUSTRATIVE EXAMPLE
Mixed Methods Study for a
Nurse Residency Program
A pre-post practice-confidence survey followed by interviews, joined in a display, for a residency program for new graduate nurses, reported to GRAMMS.
This example does not describe a real client or project. Every engagement is confidential, during the project and after it ends.
RESEARCH QUESTION
What the Study Set Out to Learn
This mixed methods study example follows a nursing education team that runs a year-long residency program for new graduate nurses. The team wants to know whether practice confidence changes over the year, and, just as important, how nurses experience the program and why confidence rises for some and not for others.
Neither question can be answered by one kind of data alone. A survey shows how much confidence changes, and interviews explain what lies behind it. The protocol is written before data collection and states why the two strands are combined, in what order, and how their findings will be brought together.
STUDY DESIGN AND DATA
A Survey, Then Interviews Sampled From It
An explanatory sequential design. About 120 nurses complete a validated practice-confidence scale at the start and at the end of the program, across several hospital units. The survey results then guide the sampling and topic guide for about 15 semi-structured interviews, drawn deliberately across the range of confidence change so that both large and small changes are heard.
- Design: explanatory sequential, quantitative strand first, qualitative strand second
- Data: paired scale scores, and interview transcripts
- Software requested: MAXQDA for coding and the joint display, with the survey analysis run in R
METHOD AND ASSUMPTIONS
Why Two Strands, and What Must Hold
The quantitative strand fixes the scoring rule before any scores are examined, including reverse-worded items and missing items, and estimates the scale's reliability in this sample. The paired change is modeled with a random effect for unit, so that nurses working on the same unit are not treated as independent, and nurses lost to follow-up are handled by multiple imputation with the observed-case result beside it. There is no comparison cohort, so the program cannot be separated from the passage of time, and the wording says so.
The qualitative strand uses reflexive thematic analysis, with coding and theme development carried out in MAXQDA and an audit trail. Quality is judged by qualitative standards such as reflexivity, coherence of themes and transparency, not by sample-size rules or statistical criteria. A joint display places the survey change beside each interviewee's account. Where nurses with large gains describe the same difficulties as nurses with small ones, those cases stay in the display as divergence and are not smoothed over.
OUTPUT AND INTERPRETATION
How the Findings Are Brought Together
The survey strand reports the mean change with its 95% confidence interval and a standardized effect size, with the reliability estimates and a respondent flow diagram. The interview strand reports a small set of themes with their definitions and illustrative extracts, with the coding decisions documented.
Integration is the core of the design, so the report states what each strand adds and where they agree, expand on one another or diverge. The meta-inferences are worded with care: the survey describes change over the year, and the interviews offer nurses' own explanations, which are plausible accounts, not tests of what caused the change. The reflexive notes record how the research team's own positions shaped the reading.
REPORTING
What the Researcher Received
A written report structured to the GRAMMS criteria for mixed methods studies, with a methods section that names the design and justifies the integration. The research question, the interpretation of the interviews and the conclusions remain with the research team.