The MedPhys match survey: Search criteria and advice for programs and applicants

Abstract Purpose The purpose of this study was to gauge the experiences of applicants and program directors (PDs) in the Medical Physics (MedPhys) Match (MPM) and to determine the most important characteristics and factors that influence decision‐making for applicants and programs when screening, interviewing, and ranking in the MPM. Opinions were also solicited from applicants and PDs on the status of medical physics residencies and the selection process, such as the availability of residency positions and satisfaction with the match process. Methods A survey was sent to all applicants registered for the 2015–2018 MPM and to all PDs registered for the 2015–2017 MPM. Survey questions asked about the pre‐interview screening, interview, and ranking stages of the residency match process. Survey data were analyzed using graphical methods and spreadsheet tools. Results An increasing percentage of applicants are female and/or hold a PhD as their highest degree. The over all number of interview invitations per applicant has increased, leading some applicants to decline interviews with the top reasons being cost of travel and scheduling conflicts. The top considerations for applicants in ranking programs were residency program/institution reputation, program structure/organization, and facilities/equipment available. The primary considerations identified by PDs for ranking applicants included impressions from the interview, personality fit, and clinical potential. While two‐thirds of applicants agreed or strongly agreed with the statement that a residency position was difficult to obtain, roughly one‐third of PDs agree that the current residency placement rate is a problem. Conclusion Four years of survey data on the experiences of applicants and PDs participating in the MPM is useful to future participants navigating the residency match system. It is hoped that the data will be helpful to inform improvements and to enhance understanding of the residency match system and how it shapes our profession.

The full survey instruments from the first two years of the survey were published as a supplement to a previous publication on ethical behaviors within the match. 7 Question types included multiple choice, select all that apply, and free response. Responses to the questions regarding opinions were collected using a 5-point Likert scale. Research Electronic Data Capture (REDCap) was used to collect and manage the study data. 8

| RESULTS AND DISCUSSION
The response rates for all surveys ranged from 28 to 33% for applicants and 48-61% for program directors, as shown in Table 1. The response rates for applicants are typical of other medical physics survey studies. The higher than normal response rate for PDs increases the validity of the data presented. According to statistics published by NMS, the percentages of registered MPM applicants who submitted a rank list were 70%, 63%, 77%, and 70% 2 compared to our survey respondent data of 84%, 83%, 92%, and 89% in the survey years 2015 through 2018 (see Table 4). The percentages of matched applicants were 27%, 32%, 37%, and 43% (official MPM demographics and including all registrants) compared to our 48%, 67%, 64%, and 43%, in the survey years of 2015 through 2018 (see Table 2). Our survey results show a bias towards respondents who submitted ranks lists and matched, as the survey matched applicant rate is consistently higher than that of the MPM demographics.
However, since the results from this survey provide important data to applicants seeking a successful match, this bias may be useful.

3.A | Applicant survey results
The demographic distribution of survey respondents is shown in

3.B | Number of applications
Applicants were asked how many applications they submitted. In the 3.C | Interviews, rank lists, and preferences Applicants were asked how many interview invitations they received and whether they declined any interviews that were offered. If a respondent indicated that they declined any interviews, they were further asked to specify the reason(s) for declining. Applicants were asked how many interviews they attended or participated in and whether the interview was on-site or conducted remotely, such as by telephone or videoconference.   Table 4 and averages to 87% over the four survey years. These values are higher than the over all percentage of applicants to the MPM who submitted a rank list (average 70%), as reported on the MPM website. 2 Respondents who indicated that they did submit a rank list were asked in a follow-up question to rate the importance of a variety of possible criteria for preferring one residency program over another. These data are shown in Fig. 4. The top attributes that respondents considered important when making ranking decisions about programs were residency program/institution reputation, residency program structure/ organization, facilities and equipment available at the institution, the work environment, and the geographic location. The least important reasons were program size in terms of number of residents and benefits packages. Table 5 shows matching statistics for all respondents and a comparison to matching rates published by MPM. Females were more likely to match than males, while imaging-only applicants were more likely to match than therapy only or than applicants that applied to both therapy and imaging residencies. The number of respondents in some categories is sometimes small and therefore challenging to draw statistically significant conclusions. A scatter plot of the number of applications submitted versus the number of on-site interviews is presented in Fig. 5, including whether or not that resulted in a successful match.

3.D | Match results and reapplication
A key question for applicants is how many interviews are needed to be successful in the match? Figure 5 may serve as a guide to help answer this question. More on-site interviews (we were unable to include the small percentage of remote final interviews) generally does result in a successful match. However, it is still possible to attend > 10 interviews and not match with a program; and it is also possible to attend < 5 interviews and successfully match. Generally, five interviews appear to be a reasonable line above which matching success is more likely. Performance during the interview is likely a key to matching success more than whether the paper application was enough to get several interview invitations. It is unclear in the survey responses what is indicated by respondents who successfully matched with zero interviews; the entry may be an inadvertent mistake on the part of the survey respondent.
Applicants were asked the total cost of attending interviews. The results are shown in Fig. 6, including a breakdown over each year of the survey. As expected, applicants who attended more interviews spent more on interviewing costs. Applicants will note that there is no correlation between the total cost spent on interviews and success in the match. It is hoped that programs will note the substantial  It is unknown from these survey data whether there is a matching advantage for an applicant to participate in a remote final interview versus in person. However, this question is now more relevant given that residency programs pivoted to entirely remote interviewing during the 2020-2021 match cycle due to the COVID-19 pandemic.
Unmatched survey respondents were asked if they intended to apply to the match in the future, as shown in Table 6. The percentages of applicants planning to reapply ("Yes") ranged from Applicants were further asked if they see the current residency placement rate as a problem for our profession. These responses are shown in Fig. 9, where there are differences in opinions expressed by male and female respondents. Survey respondents in the inaugural year of the match, particularly male respondents at 57%, suggested that the current residency placement rate is a problem.
Applicants were asked how strongly they agree with the statement "If I had known the likelihood of getting into a residency program at the time I entered graduate school, then I would have not pursued graduate education in Medical Physics." Over all responses and filtered by male and female are shown in Fig. 10.
Agreement with this statement was an alarming 39% in the initial 2015 survey (Fig. 10). While it has steadily decreased to 13% in 2018, this is clearly still an important consideration for assuring a continued supply of high-quality candidates into our profession in the future.
Applicants were asked where they believed the most appropriate place for the filter in the medical physics pipeline is. Responses are shown in Fig. 11. Note that respondents could choose more than one option. In all years of the survey, applicant respondents were   PDs were asked if they experienced changes in application numbers during their participation in the MPM. Results for all three years of the PD survey are shown in Fig. 12, where initially most programs experienced an increase in applications for the first year of the match, followed by a similar decrease in the second year of the match, and finally a relative stabilization of application numbers in the third year.

3.F | Program director survey results
PDs were also asked if their program considered or interviewed MS only, PhD only, or both types of candidates, as shown in Table 7. they offer residency positions outside of the match in addition to their participation in the match. Twelve percent of respondents indicated that they did offer at least one position outside the match. The reasons given included a starting date that did not fall between June 1 and December 31 and uncertain funding for the position.

3.G | Interviews, rank lists, and preferences
PDs were asked if they considered and if they interviewed MS only, PhD only, or both degrees in their search process (see Table 7). Zero to one PD respondent to the survey per year indicated that they only consider or interview MS degree candidates. Forty-three percent, 42%, and 28% of PDs indicated that they interview only PhD candidates. The majority of programs indicated that after considering applications, reference letters, and screening interviews, they interview both MS and PhD applicants. As shown in Fig. 13, primary considerations for interview invitation identified by PDs (with average percentage of respondents indicating this as a major factor in parentheses) included clinical potential (84%), content/quality of reference letters (80%), academic potential (73%), personality fit PDs were asked in 2017 about candidates who have not completed their graduate degree at the time of the match but were presumed to complete before the residency start date. Ninety percent of respondents indicated that they did rank applicants who had not yet completed their graduate degree, however, 78% said they seek or require assurances from the candidate that they will complete the degree by the start date, and 42% said they seek an attestation from the graduate program director or thesis advisor. The majority (60%) F I G . 9. Applicant responses to the question "Do you see the current residency placement rate as a problem for our profession?" for each year. of PDs stated that the possibility that an applicant may not complete their degree prior to the program start date is a significant consideration in ranking the applicant.
PDs were asked what their recourse would be if they matched with an applicant who does not complete their degree by the program start date. Thirty-seven percent said that they would modify the start date to allow completion of the degree, 32% said they would allow the student to begin the residency while completing the degree, and another 32% selected "Other." In 2017, PDs were asked if they considered or interviewed reapplicants during their selection process; 54% responded yes, 15% responded no, and 32% indicated that they did not know if the candidates were reapplicants. PDs who did consider or interview reapplicants were further asked what activities in the year(s) between match cycles would most increase the applicant's chance of getting an invitation to interview. Figure 15 shows the responses, where PDs were asked to choose all that apply. For programs that did consider reapplicants, additional medical physics education, research, or clinical experience (paid or volunteer) were the primary suggestions from PDs for candidates who did not match but wished to enter the match again the following year. The top two suggestions from PDs (additional medical physics education and medical physics research) correlate with the activities most likely to result in a subsequent residency match as reported by reapplicants in Fig. 7.
Visa requirements can affect applicant selection as some institutions are unable to accommodate certain types of visas. For example, US employment law only requires that a person needs to have valid work authorization. 9 PDs were asked if they accept visa applicants; 29% of respondents indicated that they accept all visa types, 46% reported that their response depends on the visa type, and 24% indicated that they do not accept visa applicants.
PDs were asked how many residency positions they offered, and responses ranged from one to twelve, with most programs offering one residency position in the match (69%, 54%, 68% in 2015, 2016, and 2017, respectively). It is acknowledged that we are not aware of any program accepting twelve new applicants in a single year, and that the answer likely instead represents the total number of residents in the program. PDs were then asked how many of those PDs were asked if they conducted remote (telephone or teleconference) interviews. Fifty percent of programs indicated that they did, with 85% using the process for screening interviews and 15% utilizing remote interviews as an option for final interviews. While on-site interviews are clearly preferred, it is anticipated that many interviews will be performed remotely in the 2020-2021 match cycle due to the COVID-19 pandemic.
Travel to participate in interviews can be costly for participants, as shown in Fig. 6. PDs were asked if their program provides financial assistance to candidates interviewing on-site. Most programs provide meals during the interview (73%), while 20% provide hotel/housing, and 2% offer funds to offset travel costs. In addition, some programs provide transportation from hotel to the airport or

3.H | Current status of residencies in medical physics
PDs were asked if they agree with the statement that there are enough residency positions available to meet current clinical demand (see Fig. 17) and where in the medical physics training pipeline is most appropriate for a filter (see Fig. 18). While applicants more strongly preferred graduate school enrollment as the appropriate filter location While this topic will not be discussed here, it is worth noting that the PD satisfaction rate from survey respondents shows a consistently decreasing trend (data presented in Interviews, Rank Lists, and Preferences section above). Both metrics related to match experience, including over all match experience and that the process needs no changes, decreased by 13% and 15% over the three years that program directors were surveyed.
Given the significant potential effects on our profession, the relative supply of and demand for medical physics residency positions is the topic of important consideration within our education and training infrastructure. 11 The number of graduates in medical physics, the number of applicants to the MPM, and the relative success of those applicants has been closely monitored, and program graduate data and match statistics are publicly available. 2,12 It is useful to compare the results from our study with these data sources.
The percentage of respondents in our survey who were accepted into a residency program is given in Table 6 and averages 62% over the four years of survey data provided. This is significantly higher than previously published estimates. 2 However, it should be noted that our data represents a small (roughly 30%), and potentially atypical, cohort of applicants. As an example, averaged over all four years, 87% of respondents to this survey submitted a rank list. In comparison, the percentage of all applicants submitting a rank list provided by the

3.I | Statistical significance
The survey data presented in the text and displayed in figures and tables in many cases represent small numbers such that no rigorous statistical significance can be claimed. Discussions of the data presented point out trends and differences in the data without the ability to estimate uncertainties. The authors have presented our interpretations and invite the reader to judge the data for themselves. Finally, we hope that concerns and suggestions identified by this survey can lead to better transparency and understanding of the match process and an improved experience for both applicants and programs.

| CONCLUSION S
A summary of takeaway points are as follows: •