RES 709 Week 7 Mixed Methods Designs Example

Bringing the Numbers and the Voices Together: A Convergent Mixed Methods Design for Hybrid Onboarding

[Student Name]

University of Phoenix

RES/709: Research Conceptualization and Design

Week 7 Assignment

[Instructor Name]

[Date]

The learner, the firms, the planned study and all figures are composites written for a model paper.

Weeks 5 and 6 designed the study's two strands: a survey linked to firm retention records and interviews with associates who stayed and associates who left. RQ3 asks what the interviews add to the survey's explanation. This paper plans how the two strands will come together, from sampling to the final interpretation, so that the study answers RQ3 rather than reporting two separate studies.

Why Mix Methods

Johnson and Onwuegbuzie (2004) argued that mixed methods research offers a third paradigm that draws on the strengths of quantitative and qualitative approaches while offsetting their weaknesses, and that pragmatism provides its philosophical grounding. For this study, the survey can estimate how strongly onboarding practices relate to retention across many associates; the interviews can reveal how and why those practices matter. Neither alone answers the problem stated in Week 4, which concerns both the size of relationships and the experiences behind them.

Core Designs

Three core designs are commonly described (Creswell & Plano Clark, 2018). A convergent design collects and analyzes quantitative and qualitative data in parallel, then merges results to compare them. An explanatory sequential design begins with quantitative data and follows with qualitative data to explain the results. An exploratory sequential design begins with qualitative data to explore a topic and uses findings to build an instrument or intervention tested quantitatively.

Choosing the Convergent Design

An explanatory sequential design would allow interview questions to follow up specific survey results, a real advantage, but would add at least six months, since interviews would wait until analysis is complete. An exploratory sequential design would suit a topic with no measures, but validated scales already exist for adjustment and embeddedness. The convergent design fits the learner's working timeline and her aim to examine the same phenomenon from two angles at once. Both strands receive equal priority, since neither research question is subordinate to the other.

Sampling Across Strands

Connecting the samples is the first integration step. All survey respondents will be asked whether they are willing to be interviewed; from those who agree, the learner will select participants to vary by retention status, in-office days and firm, so the qualitative sample reflects the range of experiences in the quantitative data. Departed associates who did not take the survey, reached through firm alumni lists, will be included to hear from people the survey might miss, with their status noted in analysis. Because the qualitative sample is small and purposeful, it is not meant to represent the population statistically; its role is to explain variation the survey reveals.

Practical Challenges

Mixed methods demands more time and skill than a single method. The learner must handle survey files, interview transcripts, separate analyses and their merger, all while working full time. Three steps reduce the load: a detailed timeline with buffers, an advisor experienced in mixed methods on the committee and a decision to limit the qualitative sample to 25 interviews. She will also take a short course in thematic analysis before data collection, since her training is mainly quantitative.

Procedure

Year 1, months 1 to 3: firms recruited; survey administered to eligible associates; retention records requested. Months 4 to 9: interviews with 20 to 25 associates selected from survey respondents who consented to follow-up, plus departed associates from firms' alumni lists. Months 10 to 15: quantitative and qualitative analyses conducted separately, each by its own methods and standards, so neither shapes the other too early. Months 16 to 18: integration, interpretation and writing.

Integration at Three Levels

Fetters et al. (2013) described integration at the design level, through the choice of design; at the methods level, through connecting, building, merging and embedding; and at the interpretation and reporting level, through narrative, data transformation and joint displays, and discussed the fit of integrated results, confirmation, expansion or discordance.

At the methods level, the two strands connect first through sampling: interviewees come from the survey sample, so their survey responses and retention status are known. At the interpretation level, a joint display will list, for each framework construct and each onboarding practice, the survey estimate beside the themes from interviews and examples of quotes, with a final column stating the meta-inference.

A row in the joint display might read: mentor assignment, odds of retention 1.7 times higher, alongside leavers saying their mentors were too busy to meet.

Interpreting Fit

Results may confirm each other, as when both show that partner contact matters. They may expand each other, as when interviews reveal why cohort time helps, through shared struggle in busy season, which the survey could not show. Or they may disagree, as when mentors relate to retention in the survey but interviews describe mentors as absent. Discordance will be investigated by examining whether the subgroups differ, whether measures captured different aspects of a practice and whether additional analysis, such as looking at mentor meeting frequency, resolves the difference.

Reporting the Results

The dissertation will report each strand's results fully in separate sections, so readers can judge them on their own terms, and then present the joint display and meta-inferences in an integration chapter. The discussion will draw on both, stating for each conclusion which strand supports it and whether they agreed. This structure lets a reader trace any claim to its evidence.

An Illustration of Expansion

Suppose the survey finds that in-person cohort days relate to retention but only modestly. Interviews might show that what mattered was not the days themselves but whether the cohort worked together on a real engagement during them. That expansion would suggest that firms should design cohort days around shared work rather than training sessions, a practical recommendation neither strand could produce alone.

Validity in Mixed Methods

Threats specific to mixing include unequal samples across strands, measures that do not correspond to themes and drawing conclusions from one strand alone. Matching interview prompts to survey constructs while allowing new themes, reporting both strands fully and stating meta-inferences explicitly address these threats.

What the Committee Will Ask

Dissertation committees often ask why a study needs mixed methods at all, given the extra work. The answer here rests on the problem statement: firm leaders lack evidence both on which practices relate to retention and on why. A survey alone would leave leaders guessing about mechanisms, and interviews alone would leave them unsure how widespread the patterns are. The design's value lies in answering both questions about the same associates, which is why the integration plan carries as much weight as either strand.

Conclusion

A convergent design with equal priority between strands, connected sampling, a joint display and explicit meta-inferences will integrate the survey and interviews into one account of how onboarding relates to retention for hybrid newcomers. Planning for discordance from the start ensures the study learns from disagreement rather than hiding it. Week 8 will assemble the full research plan.

References

Creswell, J. W., & Plano Clark, V. L. (2018). Designing and conducting mixed methods research (3rd ed.). Sage.

Fetters, M. D., Curry, L. A., & Creswell, J. W. (2013). Achieving integration in mixed methods designs: Principles and practices. Health Services Research, 48(6pt2), 2134-2156. https://doi.org/10.1111/1475-6773.12117

Johnson, R. B., & Onwuegbuzie, A. J. (2004). Mixed methods research: A research paradigm whose time has come. Educational Researcher, 33(7), 14-26. https://doi.org/10.3102/0013189X033007014