The Importance of Synthesizing Mixed-Methods Findings

While strong evaluations often utilize mixed methods, using multiple methods does not automatically strengthen evaluation (see Lyssa Wilson Becho’s 2017 blog post, Designing a Purposeful Mixed Methods Evaluation). Rather, data sources must be intentionally integrated to support interpretation. Without that integration, evaluation becomes a collection of parallel findings rather than a coherent analysis. In synthesizing evidence across methods, a coherent analysis can reveal patterns and relationships that produce insights extending beyond the reach of any individual approach.

It is important to move beyond presenting quantitative and qualitative data as separate, sequential, or side-by-side components in reporting. To capitalize on the strength of utilizing mixed-methods evaluation, careful synthesis of both components is necessary.

But really, what is the difference?

• Sequential presentation = Reporting findings from each data source separately, one after another
• Side-by-side presentation = Placing findings from different data sources together, but only describing what each data source says, without connecting them
• Synthesis = Presenting findings from different data sources together, and explaining what the data sources mean in relation to one another (e.g., how findings align, differ, or help explain one another)

A true mixed-methods synthesis examines how different forms of evidence relate to one another. Do they converge? Do they contradict? Does one help explain patterns observed in the other?

Consider this example, which shows how the same set of findings may be written up in each approach:

1. Sequential presentation

Survey results indicate that half of the participants would recommend the program and half would not. In an open-ended follow-up question, those who would recommend the program highlight the value of the skills gained, while those who would not recommend it describe logistical challenges that limited their participation. In focus groups, participants agree that the program was beneficial and that scheduling and location created barriers.

Summary: Findings suggest mixed or inconsistent participant experiences.

Recommendation: Further investigate differences in participant experience before making program changes.

2. Side-by-side presentation

Survey data show a split in whether participants would recommend the program. Open-ended responses suggest that recommendations are influenced by both skill gains and logistical challenges. Focus group participants agree that the program was beneficial but difficult to attend due to scheduling and location.

Summary: The program appears beneficial, but logistical challenges may impact participation and satisfaction.

Recommendation: Consider improving scheduling or accessibility while continuing to monitor participant outcomes.

3. Synthesis

Survey data show a split in whether participants would recommend the program. While all participants report that the program was beneficial, qualitative data indicate that logistical barriers, such as scheduling and location, limited full participation. Taken together, these findings suggest that differences in recommendations are driven not by disagreement about program value but by how participants weigh the benefits of skill development against the challenges of participation.

Summary: Participants consistently value the program, but the split in recommendations reflects differences in how participants weigh benefits versus feasibility. All participants experienced both strong outcomes and logistical barriers; the difference lies in what they prioritize.

 Recommendation: Maintain core program content while addressing logistical barriers specifically for participants with limited time or travel flexibility, as improving feasibility may increase endorsement without changing program design.

• Presented sequentially or side-by-side, these findings may appear contradictory or inconclusive. The survey suggests a divided experience, while the focus group suggests broad agreement that the program was both valuable and logistically challenging. A sequential presentation leaves the reader to reconcile this tension independently, and a side-by-side presentation may yield partial insight (e.g., logistics are a general barrier to engagement). However, without explicit integration, the likely interpretation is that participant experiences were inconsistent or that the program worked for some but not others. This can lead to vague or unfocused recommendations, as the underlying reason for the split is unclear.
• Synthesized, a different interpretation emerges. The split in recommendation is not about disagreement over program quality; it reflects differences in what participants prioritize. All participants experienced both benefits and burdens. Those who prioritize skill development recommend the program despite logistical challenges, while those who prioritize feasibility do not. The focus group clarifies that the divide is not about whether the program “worked,” but about how participants weigh tradeoffs. This presentation yields a precise insight (e.g., the issue is not logistics alone, but how different participants respond to the same conditions), and the resulting interpretation points to a specific and actionable insight.

Still, mixed-methods integration is not just about combining quantitative and qualitative data. It’s about integrating perspectives across stakeholders, roles, and experiences.

As in this example, a thoughtful synthesis of mixed-methods data can support more comprehensive, actionable recommendations. To move from side-by-side or sequential reporting to true synthesis, consider these strategies:

• Organize findings by evaluation question, not by method or respondent.
• Explicitly compare data sources within each section. Note where findings converge and where they diverge.
• Use explanatory language, such as “These interview findings help explain…” or “While survey results suggest…, qualitative data indicate…”
• Highlight implications that emerge from integration, not from a single data source alone.