Design

A parallel convergent methods design was used in this study (Fig. 2), involving quantitative (cross-sectional study) and qualitative (framework analysis methodology) approaches for data collection in the same period (January 2022–July 2023) and the same study area (Guangdong Province, China). The quantitative and qualitative research strands were conducted and analyzed independently, and parallel participants were chosen for enrollment in the study to avoid the limitations of repeated measurements and cross-sectionality of results due to inclusion of the same participants both in the quantitative and qualitative components. Quantitative and qualitative data were integrated in the overall interpretation.

Fig. 2

Flowchart describing the research method

Quantitative component

In the quantitative phase, a descriptive cross-sectional study was conducted to explore factors affecting fertility intention to among two-child mothers. The quantitative phase of this study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) [21] reporting guidelines.

Sample and setting

Participants were recruited into this study from Guangdong Province through convenience sampling through the community networking groups on WeChat software. Eligibility criteria for participants include the following inclusion criteria: ① age range of 20–49 years; ② married with two children (including twins) [13]; and ③ consent to participate. Participants were excluded from the survey if they ① had a history of mental illness or any cognitive disability or ② were pregnant (given that psychological health can be affected by changes in hormone levels).

According to the empirical method, the sample size of a regional social research study needs to be greater than 500 [22]. A total of 631 participants were enrolled in the study, and 28 invalid questionnaires were excluded because of incomplete or contradictory responses. Finally, 603 valid questionnaires were used in the final analysis.

Measurements and instruments

Quantitative data were collected with the following three instruments.

1.

Sociodemographic characteristic

A self-designed questionnaire was used to investigate and collect the sociodemographic characteristics through 12 questions pertaining to age, highest education level, monthly family income, occupation, children’s sex and main caregivers, maternity insurance, old-age insurance, whether participants and their husbands were singleton, and the children’s grandparents’ physical health status and life support.

2.

Determinants of the fertility intention scale

The fertility intentions scale was designed by Zhou [23] to explore the factors affecting fertility intention. This scale contains 32 items, assessing eight dimensions: support in the form of available resources, social support, policy support, perceived values, emotional values, perceived risk, perceived behavioral control, and attitudes. After obtaining copyright permission for the scale, we modified several expressions to reflect the purpose of this research. For instance, “fertility risk has a significant impact on fertility intention” was modified to “fertility risk has a significant impact on third-child fertility intention,” thus providing a more targeted understanding of the third-child fertility intention. This multidimensional scale was scored on a Likert scale of 1 to 5 (“strongly agree” to “strongly disagree”). In our study, this fertility intention scale has good internal consistency (Cronbach’s α = 0.818).

3.

Measurement of fertility intention

We used the question: “Do you have the intention to have a third child?” to explore the third-child fertility intention. The response options were “no intention at all,” “no intention,” “uncertainty,” “intention,” and “strong intention.” “No intention at all” and “no intention” were defined as no desire to have a third child, whereas “intention” and “strong intention” were defined as a desire to have a third child. The intention to have a third child served as the dependent variable and the primary outcome.

Data collection

We Chat was the main venue for disseminating recruitment announcements for population selection. The Questionnaire Star website was used to disseminate questionnaires, because it is the most popular social media platform in China, with 1.15 billion active users. Second-born mothers of childbearing age living in Guangdong Province were invited to fill out the questionnaire. We compiled a standardized set of instructions including the purpose, significance, and content of the survey. Participants were informed that their participation was anonymous and voluntary. To ensure effective data collection, all survey items were required to be answered before the form could be submitted. Participants were assigned identification numbers, each of which could be used only once. After the data collection, the questionnaire responses were collated and examined. Questionnaires were excluded if ① responses to at least one-third of the total questions were lacking; ② only one answer choice was chosen across the entire questionnaire; and ③ answers were not filled in as instructed.

Statistical analyses

Data analysis was conducted in SPSS 27. 0 software (IBM Corp., Armonk, NY, USA). We used Q-Q plots to test the normality of the data, thus confirming a normal distribution. In the descriptive analysis, frequencies and proportions were calculated for categorical and ranked variables. For continuous variables, we calculated means and standard deviations (SDs). Univariate analysis was performed with the chi-square test and the Kruskal–Wallis test for categorical and ranked variables. Ordinal logistic regression analysis was performed to investigate the factors of fertility intention, and the odds ratio (OR) and corresponding 95% confidence interval (CI) were calculated. Variables with statistical significance of 0.25 or less in the univariate analysis were entered into the logistic regression model. All statistical analyses were two-tailed, and the criterion for statistical significance was set at 5%.

Qualitative component

Qualitative research is a branch of science, which devoted to the understanding of social phenomena from the perspective of the subjects and to describe their cognition and thoughts in the way that they have experienced [24]. Perceptions and views of participants on the third child fertility intention are worth exploring to complement the results of quantitative research. Hence, on the basis of framework analysis methodology, a semi-structured interview was conducted to investigate the perceptions and consideration of mothers of childbearing age with two children regarding the willing to have another baby. Reporting followed the Consolidated Criteria for Reporting Qualitative Research (COREQ) [25] guidelines.

Measurements and instruments

A preliminary interview outline was developed by the research team according to analysis of the available studies on this topic and the conceptual model. Subsequently, we conducted two rounds of consultations with five experts with master’s degrees or above, and senior professional status. All experts had more than 10 years’ work or research experience in the fields of obstetrics nursing, psychology nursing, sociology, or public health medicine. The effective response rate in both rounds of expert consultations was 100%, and the advice presentation rate was 80% in the first round and 40% in the second round. The authority coefficient of the experts ranged from 0.86 to 0.88, thus indicating that the results of the expert consultations were highly credible. Subsequently, the pre-interview outline was formulated primarily on the basis of the consultations. Next, two mothers of childbearing age with two children who met the inclusion and exclusion criteria were selected for pre-interviews. After several modifications were made according to the results of these pre-interviews, the final outline of the interview script, including nine items, was created (Appendix A).

Data collection

Data were collected through individual semi-structured face-to-face interviews conducted in the Chinese language at the time and place as agreed upon by the two parties. First, we hid the information of each participant and pseudonymized by number. Subsequently, the interview was conducted in a quiet setting to ensure confidentiality and privacy (e.g., in a secluded garden). Two interviewers who had not had any contact with the participants before the study conducted the interviews. One interviewer conducted an in-depth interview with the participants by referring to the interview outline, and the other was responsible for recording the required details and re-questioning when necessary. To establish trust and effective communication with the participants, the interviewer posed several general questions at the beginning of the interview. Next, according to the interview outline, the interviewer commenced the interviews by referring to topics associated with the fertility willingness. Each topic was introduced with an open-ended question before follow-up questions were asked, and prompts were used when needed. Each interview lasted approximately 20–30 min. All semi-structured interviews were audio-recorded and transcribed verbatim in full in Chinese. Field notes provided important details, such as the atmosphere and participants’ non-verbal behaviors and emotions, thereby enriching the data analyses. After the eleventh interview, data saturation was reached, because no new themes were identified. Consequently, the data collection process was terminated.

Statistical analysis

The framework method was used to analyze the semi-structured interview data. In the first and second phase, after the interview, audio recordings were transcribed verbatim by trained research assistants within 24 h and reviewed by the second researcher to ensure content accuracy. After transcription, the transcriptions were reconfirmed by the study participants to ensure that the transcription text matched the content of the interviews and familiarisation with the interview. In the third and fourth phase, on the basis of the conceptual model, we formulated general meanings for significant statements until a consensus was reached. All the interview data were coded with Nvivo12plus. Significant statements and high-frequency words were refined. We arranged these statement meanings into clusters of themes to identify common significant concepts, which were prototypes of themes. In the fifth and sixth phases, the research team conducted a brainstorming session to detail early themes by further refining significant concepts and comparing similar themes to extract common viewpoints, which then formed the emerging themes. Subsequently, these emerging themes were organized into major themes and subthemes, and three or four typical statements were included for each, to provide support. In the final phase, all themes were discussed among the research team and modified if required.

Mixed methods integration

The quantitative data obtained from the questionnaires and the categories derived from the inductive qualitative analysis were compared through side-by-side comparisons and joint displays, to merge them visually, and to elaborate upon how they complemented each other, thereby generating additional inferences [26].

By integrating the findings of both methods, we developed deeper insights and a better understanding of the various factors that impact the intention on the third child. Given the exploratory nature of this study, multivariate results were considered during the mixed methods integration for quantitative findings.

Ethical considerations

This study was approved by the Ethics Board of Guangzhou Medical University in Guangzhou City, China (approval number 202201001). All participants recruited for this study, in both the quantitative component and the qualitative component, provided informed consent. Respondents were assured that they could cease participation in the research at any time and would not experience any consequences. Their responses and personal information were kept confidential and used only for research and data analysis.