“I was told to find what broke my heart and fix it:” College students explain why they want to become teachers

Abstract Using the open-ended question from the FIT-Choice scale, this study presents an examination of why college students choose teaching as a career, in their own words. Over 100 responses were analyzed with two coding systems: thematic coding from the text and a coding system generated from the FIT-Choice survey questions. Two raters reliably coded independently. Across the two coding systems, students reported choosing teaching as a career because they want to teach and want to help children. They were motivated by the ability to make a difference in the world through teaching. The coding system based on FIT-Choice provided similar results to the students’ responses to the quantitative FIT-Choice measure, with social utility value and intrinsic career value being ranked highest. Open-ended responses support the survey responses and suggest that these pre-service teachers are choosing teaching for altruistic reasons.


Subjects: Teacher Education & Training; Initial Teacher Training; Teacher Training; Teachers & Teacher Education; EducationalPsychology
Jennifer D. Moss ABOUT THE AUTHOR Jennifer D. Moss, PhD, is currently an assistant professor of psychology at Emporia State University, in Emporia, Kansas. She earned her PhD in educational psychology at Purdue University, where this work was conducted. Her research focuses on helping instructors, from preschool teachers to college professors, create environments where both they and their students thrive. Her publications focus on teacher motivation within a self-determination theory context.

PUBLIC INTEREST STATEMENT
This paper explores college students' statements on why they want to become teachers and compares these statements to survey results. Researchers have been studying why college students want to become teachers for decades and have seen a variety of trending reasons. Looking through two different lenses, these recent responses suggest that today's future teachers are choosing the profession for altruistic reasons-wanting to teach and wanting to help children. The short-answer, open-ended responses the students provided agree with their survey responses, demonstrating that whether asked directly or indirectly, students choose teaching for positive motives.
In a time when the teaching profession is being heavily politicized and teachers' ratings of job satisfaction are falling, it is encouraging to hear directly from future teachers that they are choosing the profession to help children and change the world. As we continue our work with preservice teachers, it will be useful to remember what is motivating them and to build upon this positivity in our programs and our classrooms.

Introduction
Teaching is often characterized as a noble profession, one in which many sacrifices are made for few wages. As one student stated when asked why she was choosing to become a teacher, "I was told to find what broke my heart and fix it." However, teachers also describe being put-upon by increased testing and paperwork requirements, demanding parents, and students who are less and less ready for school. Today's political climate provides many reasons why a college student might choose not to become a teacher: threat of potential layoffs, excessive time spent on testing, and the negative way politicians portray teachers and other public servants (Layton, 2015). Despite declines in enrollment in teacher preparation programs (Sawchuk, 2014), despite fears of potential layoffs (Theisen-Homer & Levinson, 2017), and despite discourse from some politicians that may be seen as anti-teacher (Layton, 2015), students still enter colleges of education to earn their degrees and their teaching certificates. Are these students pursuing a teaching certificate as a fallback career, in case their first choice does not pan out? Are these students entering the profession out of a sense of duty or pressure from others? Or are they interested in teaching for altruistic reasons, seeking to make a difference in the world?
In order to appreciate the decisions of collegiate pre-service teachers, students were asked to share their reasons for entering the profession. As a part of a larger study, students enrolled in three different initial teacher-preparation courses at a large Midwestern university were asked to complete the following question: Please briefly state your main reason(s) for choosing to become a teacher. This question originates from the FIT-Choice (Watt & Richardson, 2007) assessment that was designed to better understand why students choose teaching as a career. The FIT-Choice includes this one open-ended question as well as 50+ questions that students answer on a seven point Likert scale with their level of agreement or disagreement. Since the participants in the study not only completed the open-ended question but also the Likert-scale questions on the FIT-Choice, overall comparisons between the open-ended and survey responses can be made. Importantly, this is the first study to assess the open-ended question that is part of the FIT-Choice measure. The authors of the measure have been collecting the data since they began using the measure (Richardson & Watt, 2006), but have not formally analyzed the open-ended questions (H.M.G. Watt, Personal communication, 2015;H.M.G. Watt, Personal Communication, 2018).

Why college students choose teaching
There is a large body of literature detailing why college students choose a career in teaching. In the mid-1950 s, Fielstra (1955) found that students were influenced by friends and family who thought they should be teachers, along with accounts in the media that suggested there was a need for teachers. These students were not influenced by the money that they would (or would not) make in the profession, but they cited the short work-weeks and the summers off as contributing reasons for choosing teaching. Since the time of Fielstra's work, researchers have reported that college students have chosen to become teachers for a variety of reasons, from altruism (Brookhart & Freeman, 1992) to social influence (Schutz, Crowder, & White, 2001), to enjoying the subject they will teach (Kyriacou, Hultgren, & Stephens, 1999) all the way to students who reported choosing teaching as a fallback career (Haubrich, 1960).
Following several cohorts of United Kingdom teachers from 1960 to 1990, there was a significant drop in how many college students chose education as a field of study (Chevalier, Dolton, & McIntosh, 2007). At that time, there was a deficit of 34,000 teachers in the UK. Watt and Richardson (2007) who developed the Factors Influencing Teaching Choice, or FIT-Choice, also focused their work on deciphering teacher shortages by identifying what motivates students' interest in the career. Their framework provides a fine-grained analysis of what motivates preservice teachers, or college students pursuing a teaching certification, to choose a career in education.
The current climate for teachers in the United States could be described as poor. Teachers have become political pawns in elections and budget crises (Swalwell & Schweber, 2016). They report that the curricula they are required to teach consists of administering practice tests in preparation to take standardized tests (Turner, 2017). Job satisfaction among teachers has been declining steadily. In 2008, 62% of teachers reporting being satisfied, a number that dropped to only 39% reporting satisfaction in 2012 (MetLife, 2013). Former New Jersey Governor Chris Christie went so far as to suggest that teachers' unions deserve a punch in the face for being the "single most destructive force in public education," (Layton, 2015). As a final example, Wisconsin Governor Scott Walker, a 2016 presidential candidate, said that since he had shown how he could take on teachers and other public workers who were protesting budget cuts, he could take on global terrorism (Strauss, 2015). The journal New Educator devoted a recent issue to preparing future teachers in during the Trump Presidency (Blumenreich, Baecher, Epstein, & Horwitz, 2018) describing that they received an unprecedented number of submissions from teacher educators who want to share their thoughts on issues including the ongoing history of racism and homophobia in schools, working with undocumented students, and how to help young children deal with the social anxiety of the times. Perhaps it should be no surprise that enrollments in colleges of education around the country have been falling steadily since 2010 (Berg-Jacobson & Lindsay, 2016;Sawchuk, 2014;Sobota & Coulter, 2013). A study released in December of 2019 indicated that enrollment in teacher preparation programs has dropped by one-third between 2010 and 2018, while all bachelor's-level college enrollment has increased during the same time period. The one-third drop in teacher program enrollment represents the equivalent of 340,000 students who have chosen to earn a college degree in a non-teaching field (Partelow, 2019).

Theoretical and methodological frames
In this study, there are two theoretical frames used for the analyses. First, grounded theory was employed to identify themes directly from the text of the participants' responses. Grounded theory (Charmaz & Smith, 2003) is a flexible method for extracting themes from your data. Inductive grounded theory allows researchers to approach their data without preconceived notions, and develop codes to use when working with the data from their reading of the texts (Bernard, 2006). Inductive grounded theory was used to develop the thematic codes from the text.
Secondly, expectancy-value theory (Wigfield & Eccles, 2000) forms the foundation for the FIT-Choice measure. According to Watt and Richardson (2007) the authors of the FIT-Choice framework, key reasons why people choose to undertake an activity or an academic subject are their belief or expectancy of success and their value for the task. The FIT-Choice framework allows researchers to identify the values that preservice teachers hold for different aspects of a teaching career (Richardson & Watt, 2006). In the current study, the FIT-Choice framework was used to create a coding scheme for the open-ended question, and for comparison of the survey responses to the open-ended responses.
This research project is important for two reasons. First, it represents a unique opportunity to study the open-ended responses of over 100 students, a large number in a qualitative study. Frequently, qualitative work focuses on depth, not breadth, with small numbers of participants (e.g., Schutz et al., 2001;Whitbeck, 2000). However, because the open-ended question was part of a survey that was administered to many students, a large number of responses were available for analysis, allowing for robust analyses. It is worth noting that in Brookhart and Freeman's (1992) seminal review of published and unpublished studies from 1960 to 1990 examining why college students choose a career in teaching, only three studies used student writing as the primary source of information. Regarding sample size in these three studies, two were "not given," and one was 91 students. In an updated review, Heinz (2015) found only four out of 46 studies from 1992 to 2014 that examined qualitative questionnaires, with only two of these having a sample size over 101 students. As there have been few studies examining large numbers of student responses, the current study, with 112 responses, is very meaningful.
Secondly, the open-ended responses from the FIT-Choice measure have not yet been the subject of published research. As of now, researchers' attention has focused on the survey/quantitative data (e.g., Fokkens-Bruinsma & Canrinus, 2012;Glutsch & König, 2019;Hennessy & Lynch, 2017;. Working with both sets of responses will generate a robust picture of students' motivation for a teaching career, and provide the ability to compare the open-ended answers to the quantitative FIT-Choice responses to see if both sets are consistent or if open-ended statements differ from survey responses. The first goal of this study was to analyze what teacher-education students say when they are asked an open-ended question about why they are choosing a career in teaching. The second goal was to compare students' responses to the open-ended question and a survey on the same topic. In order to accomplish these goals research has been initiated with the open-ended question from the FIT-Choice measure (Watt & Richardson, 2007) which asks students to explain, in their own words, why they want to become teachers. The students' responses were analyzed with two coding schemes to look for common themes, and then the open-ended answers were compared to the FIT-Choice survey responses to examine potential similarities.

Participants
Students were recruited from 3 of the 5 introductory teacher preparation courses at a large, Midwestern R-1 university. The other two introductory courses were not used for recruiting purposes because of significant overlap of enrolled students. Institutional review board approval was obtained before any recruitment or study activity began. A short presentation was given at the end of class and students returned a form indicating that they were interested in participating. Those who indicated interest were sent the survey by email, which began by asking for the students' consent and included the open-ended question. Of the 219 students who indicated initial interest, 112 students consented and participated in the study. In general the students were White, women, not Latino/a, and in their second year of college. The racial breakdown was as follows: White, n = 103; African-American, n = 2; Asian, n = 3; and no response, n = 1. Ethnicity included Latino/a, n = 7 and not Latino/a, n = 105. There were 96 women and 16 men in the sample. The students' class standings were freshmen, n = 21; sophomores, n = 66; juniors, n = 16; seniors, n = 6; and unidentified, n = 3. Fifty-eight students indicated that they were planning to teach elementary school, 34 students said they planned to teach high school, and the remaining 20 students did not indicate a choice. The demographics of the participants were in line with the demographics of the college of education at the university. It was expected that most students would be report being sophomores as their class standing because the courses being surveyed were designed to be taken in the students' second year of the teacher preparation program. Differences were not noted among the different ethnic, racial, gender, or class-standing groups. Differences were also not noted between the students planning to teach elementary or middle/ high school. Participating students were given a 5 USD coffee shop gift card as a thank-you gift.

Measures
The main measure in this study is the open-ended question from the FIT-Choice (Watt & Richardson, 2007) framework. The text of the question asks students to "Please briefly state your main reason(s) for choosing to become a teacher." It is important to note that students were not influenced by the FIT-Choice survey questions in responding to the open-ended question. The open-ended question is the first question of the FIT-Choice survey and comes at the beginning, before other potential reasons for choosing teaching are shown to the participant.
In addition, the open-ended responses will be compared to the survey responses from the FIT-Choice framework (Watt & Richardson, 2007), the other formal measure included in the study. In the administration of the FIT-Choice, the survey is presented after the open-ended question, and, using an expectancy-value perspective (Wigfield & Eccles, 2000), asks students about what has influenced their interest in choosing a career in teaching, what their beliefs are about teaching, and about their decision to become a teacher. Using a 7 point Likert-scale, students are asked to rate the influence of particular statements on their decision to become a teacher, including, "I have the qualities of a good teacher," "As a teacher, I will have lengthy holidays," and "Teachers make a worthwhile social contribution."

Design/procedure
The 112 open-ended responses were analyzed using content analysis, a method that is used in qualitative and quantitative research. Simply put, content analysis is the act of describing the meaning of a particular text (Schreier, 2012). Texts can be lengthy, such as entire books, or short, such as the responses in this study. In support of using short statements, it has been said that shorter passages are more focused and lengthier passages provide more variable data (Weber, 1990). In the case of the responses that were obtained on the open-ended question in the current study, the unit of analysis is the entire statement, because the statements are brief. Longer passages would require divisions for specific units of analysis (Bernard, 2006). In this study, the responses to the question ranged from two words to 48 words, with a median number of words of 15.4.
Two frameworks/theories were used in the coding of these open-ended responses. As was referenced earlier in the paper, grounded theory (Charmaz & Smith, 2003) was used to create the thematic codes from the text. Specifically, inductive grounded theory (Bernard, 2006) allowed the raters to examine the texts and identify themes without preconceived ideas. Selected factors from the FIT-Choice survey measure (Richardson & Watt, 2006) formed the basis for the FIT-Choice coding system.
Prior to the beginning of coding, a codebook was created. This codebook describes the codes for each system and which types of statements would be exemplars of that particular code. The codebook also explained the process for coding when the statement appears to be referencing two or more codes. In the sample of 112 responses, only 20 responses could have been coded into multiple categories. The second rater completed a one-hour training session with the author of the study in order to understand the different coding schemes and the specific codes for each scheme. The author and the second rater also reviewed and agreed upon the hierarchy of coding for each coding scheme. The codebook and the two spreadsheet files with the student responses, repeated for each coding scheme, were provided by the author to the second coder. Ten responses were coded initially, and the author and the second coder calibrated their ratings. After that, the remaining responses were coded. Upon completion, reliability of each coding scheme was calculated, and then the author and second coder came to consensus.

Thematic coding
Four hierarchical themes arose from the data: like children, want to help children, want to teach children, and inspired by other teachers. These thematic codes were designed to be mutually exclusive so that accurate comparisons of why the students chose teaching could be made. If responses were allowed to be coded with multiple codes then the data would become indistinguishable and muddy. By allowing responses to be coded into only one category, this allowed the clarification of the data into the strongest reasons for choosing teaching. A simple statement that would be easily coded into one category is, "To make a difference in the lives of young people." This statement was coded in the want to help children category. A statement where the hierarchy would be used is, "I want to be a teacher, and I want to stay with children." In this second statement, the student references both a desire to teach and how she likes and wants to continue working with children. In this case, the statement was coded at the want to teach children level. Table A1 contains the four themes that were discovered from the data.
If someone is choosing teaching as career, she would at least like children. The next level would be having a desire to help children, which includes that she also likes children. Moving beyond those steps, she should like children, want to help children, and want to take on the role of being a teacher. If she is interested in taking on the role of being a teacher, she might also say that she was influenced by a previous teacher. Each statement level supersedes the previous one and encompasses the prior motivations as well. If the statement did not fit the classifying system at all, it was labeled unclassified. The hierarchy, presented in Table A2, was only called upon if a student had written a statement that contains multiple themes. Raters read the items, and chose the theme that offered the best fit, referring to the hierarchy if there were multiple themes in the answer.

FIT-choice derived coding
Since the open-ended question comes from the FIT-Choice measure (Watt & Richardson, 2007) a coding scheme derived from FIT-Choice was also created. It is a goal of this research to determine how well the responses meshed with the participants' answers on the quantitative portion of the FIT-Choice measure.
The FIT-Choice survey questions are grouped into many factors. The factors that were used for coding the open-ended responses included personal utility value, social utility value, and intrinsic career value. Other FIT-Choice factors, which were not used as part of this coding scheme, tapped into ideas such as belief that teaching is hard work, or that teaching requires specialized knowledge. These factors were not used as they do not answer, as clearly, the same question as the open-ended question regarding why a student wants to be a teacher. The descriptions for "how to code" using these categories were taken from the text of the FIT-Choice questions which loaded onto the respective factors. See Table A3 for the FIT-Choice factors used in this study and sample survey questions for each factor.
The raters then read the items, and referred to the hierarchy in Table A4 if there were responses that could be coded in more than one category. As with the thematic coding, both raters discussed the hierarchy prior to coding and agreed upon the order. The hierarchy was only used when an item referenced two categories. A sample statement that was easily coded into one category, prior teaching and learning, is "An experience I had while in school." A statement where the hierarchy was used to clarify the category is, "I had many teachers that invested in me and I love to watch people learn and grow." This statement could have been coded into prior teaching and learning, but ultimately using the hierarchy, was coded as intrinsic career value because the student mentions her love of watching people grow and learn. Table A4 provides an explanation for the order in the hierarchy.

Reliability
Once inter-rater reliability was calculated based upon the raters' initial, independent codes, the raters met and came to consensus. The consensus codes are the ones reported for analyses. The reliability statistics are reported below for each coding system.

Thematic coding
The reliabilities for the thematic coding were the higher of the two coding systems. Inter-rater reliability was calculated for each coding system with a kappa statistic. For kappa, values greater than.60 are considered adequate and kappa is used to overcome chance agreement (McHugh, 2012). For the thematic coding the kappa was equal to.88 which is considered high. This high level of reliability may have been due to the fact that there were only five categories, allowing the raters to differentiate among the responses, without reducing the reliability with more categories.
As the themes arose from the texts, all of the five categories were used during the coding of the responses. Table A5 shows the distribution of the coded responses, along with samples from each category. In this case, close to half of the students, 44%, said that they want to teach. This was the highest level in the hierarchy and, from the thematic coding, these students were the ones who indicated that they were truly interested in providing education and managing aspects of the classroom. A slightly smaller number, 38%, were interested in helping children. These students indicated that they want to be of service to children but did not specifically mention school or classrooms. Students who were inspired by other teachers constitute 13% of the sample, and students who like children, 6%. The three unclassified answers represent 3% of the sample.

FIT-choice coding
The reliability for the FIT-Choice coding, derived from Watt and Richardson's FIT-Choice measure (2007), was also found to be more than adequate. The kappa was equal to.66. The slightly lower result of the kappa from the FIT-Choice coding, compared to the thematic coding, may be attributed to the larger number of categories that the items could be coded into. There were eight available codes for the FIT-Choice derived coding system. Table A6 shows which codes were used, how many responses were coded into each category, and sample responses from each group that was used. It is worth noting at the beginning of the discussion that no student responses were coded into the fallback career category. An overwhelming percentage of student responses, 66%, were coded into the social utility value category. Intrinsic career value was the second highest category, with 24% of the responses coded in this category.

Comparisons between the two open-ended coding schemes
A difference that is apparent when comparing the thematic coding and the FIT-Choice derived coding is the placement of statements about the influence of prior teachers. We coded the responses thematically first, without the overlay of the FIT-Choice framework. This led us to place the prior teachers' influence at the highest spot because those who said they wanted to be a teacher due to the influence of prior teachers were thought to be more likely to be entering the profession with a clear understanding of the demands. However, when we examined the responses through the lens of the FIT-Choice and expectancy-value theory, we viewed the influence of prior teachers as more similar to responses referencing ability, where students were answering that they chose teaching because it was something they thought they could do, and to responses of social influences, where they chose teaching because they have seen someone do it.

Comparisons between fit-choice open-ended coding and surveys
The third goal of this study was to compare the open-ended coding that was derived from the FIT-Choice measure (Richardson & Watt, 2006) with the overall results obtained from the FIT-Choice survey measure that students also completed. This was done to compare whether the students reported similar answers at the beginning of the survey on the open-ended question as they did during the phase with the 54 survey questions. At the initial data collection, students answered the open-ended question before beginning the survey measure, as the FIT-Choice protocol dictated.
Comparing the open-ended and the survey answers side-by-side is useful and enlightening. It is important to remember that for the open-ended coding, the raters were limited to one category; any given answer might have had multiple reasons, but was only coded by the highest point on the hierarchy. For the survey, the students were asked to respond to 54 items which they ranked on a Likert-type scale from 1 as not at all important to 7 as extremely important. The survey then collected students' thoughts about all of the potential reasons for choosing teaching, rather than coding their answers into one category.
In order to compare the answers side by side, Table A7 shows the categories and the rank order for each measure. In the case of the open-ended measure, the order was derived from the number of participants whose answer was coded in that category. In the case of the survey measure, the number represents the order of the mean of the responses for questions that fit that category. For example, the category of Social Utility Value, the largest number of open-ended responses fell into this category, while that same category had the second largest mean score when the same students took the Likert-type question. Table A7 it can be seen that the open-ended answers are well-aligned with the responses to the survey. In their open-ended responses, students described social utility value and intrinsic career value as their two most significant influences in choosing to become a teacher. Using the survey responses, students rated intrinsic career value highest and prior teaching and learning second. The remainder of the rankings are very similar, however, fewer than 10% of the openended responses were coded in the categories of prior teaching and learning, social influences, ability, and personal utility value with no students responding with a fallback career response. While the hierarchy imposed on the open-ended questions may be contributing to this shift in the data toward social utility and intrinsic career value, it is also plausible that students respond differently to the openended questions and the survey questions. For example, ability was only coded by one respondent as why she would choose a career in teaching. However, the mean score was over 6 on the 7 point survey scale, indicating that students were highly endorsing the influence of ability in their decisions to become teachers. The differences in responses may be due to social desirability bias where students felt the need to give more preferred-sounding answers to the question in an open-ended context. Unfortunately, due to concerns of survey fatigue, social desirability was not measured in this study.

Discussion
This study has two goals: first, to analyze what teacher-education students write when they are asked why they are choosing a career in education; second, to compare the analyses of these open-ended responses to survey responses on the same topic.
The first goal of analyzing the students' open-ended responses was met. Responses from 112 pre-service teachers were read carefully and a codebook for two different coding systems was created. Answers were coded and each coding system was examined to see which categories received the most codes. The preservice teachers' voices were clearly heard through their responses. These students are a group that values teaching for the way it can improve society. They have said that they want to teach and want to help children.
In order to fully analyze the students' responses, the next step was to look across the two coding systems for common themes. Two common themes were identified in the open-ended responses. First, according to their statements, prior teachers did not significantly influence these students to enter the profession. The categories "inspired by other teachers," from the thematic coding, and "previous teaching and learning experiences" from the FIT-Choice coding attracted few responses. Only four responses were coded in this category in the FIT-Choice scheme, and only 15 were coded similarly in the thematic coding out of 112 students.
Measures that examine why students become teachers often ask about the influence of prior teachers (Sinclair, Dowson, & McInerney, 2006;Thomson, Turner, & Nietfeld, 2012). It is noteworthy that the levels reported by the college students in this present study in the open-ended responses are lower than other prior research, and lower than the same students' survey responses. Reaching back to a 1990 study (Hayes, 1990), more than half of the students, 54%, said that they were influenced to choose the profession by previous teachers. Watt and Richardson (2007) the authors of the FIT-Choice measure, reported in a study of two preservice teacher populations that the reported influence of prior teachers was notable, at 5.09 and 5.07 out of 7. Irish preservice teachers who completed the FIT-Choice survey reported prior teaching and learning experiences as their highest motivation with a mean of 5.57 of 7 (Hennessy & Lynch, 2017). The students in the present study reported a mean of 5.5 on the survey measure indicating that they were very strongly influenced by previous teachers in their decision to become teachers. While the number of respondents whose answers were coded in this category was small on the open-ended measure, the same students reported a high mean value for the same construct. This discrepancy may be due to the open-ended question coming first and not providing any prompts or cues about why the respondents might want to be a teacher. The survey provides options and students are asked to rank those options. When presented with the options, perhaps the students then identified that they were influenced by a former instructor.
The second common theme looking across the two coding schemes is that this group of preservice teachers is making the choice to teach for altruistic reasons. From the lens of the FIT-Choice derived coding, the results indicated that students were choosing teaching for the social utility value. Their responses included ways that they could improve society such as helping special needs students and teaching children to think for themselves. Coding of the responses with the thematic coding scheme demonstrated that 87 of the 112 students were choosing teaching either because they want to teach or that they want to help children. Taken together, these are very altruistic reasons for choosing a difficult career during a difficult time. As was discussed previously, politicians are demeaning teachers, and while the amounts of paperwork and testing increase, teachers feel that they are losing control of their classrooms. However, the students in this study are still choosing to teach. It is worth considering whether the climate that downgrades teaching is making students more stalwart in their interest in teaching or if these remaining students are the stalwart preservice teachers that would have enrolled in colleges of education in any climate. As was also noted previously, enrollment in teacher preparation programs has decreased in previous years. Perhaps if there were not a drop in enrollment, numbers of students choosing teaching for less adaptive reasons would be seen. Since the climate for teachers is not as positive as it has been previously, it may be that the students who would have provided answers in the fallback and personal utility value categories are choosing different college majors now.
The second goal of this paper is to compare the FIT-Choice open-ended coding scheme which was developed for this study to the original FIT-Choice survey. The majority of the categories were similar in the order of responses and mean answer scores. As this is an exploratory study, further work will be required to why students reported social utility value the most often in the openended responses and intrinsic career value had the highest mean of all the survey categories. Additionally, further research will be able to determine if ability is consistently as low in the openended responses and high in the survey as it was in this sample.

Limitations and future considerations
This paper does have a key limitation: the students are all from one university's teacher education program. This limits the ability to generalize the findings beyond these students, even though the sample size of 112 was sufficient to achieve saturation (Creswell, 1998). Results from other studies using the FIT-Choice survey (Glutsch & König, 2019;Hennessy & Lynch, 2017), along with studies that used different measures (Al-Yaseen, 2018;Lee & Kim, 2018;Tomšik & Gatial, 2018) suggest that the responses of the students are very context dependent, with variation reported in whether intrinsic or extrinsic motives were more salient or whether ability was a key determinant or not when choosing teaching as a career. Future work examining the open-ended FIT-Choice question along with the survey results should include much larger samples of students across multiple universities in order to enhance generalizability.

Conclusion
This initial examination of the open-ended question on the FIT-Choice scale (Watt & Richardson, 2007) has added to the work in preservice teacher motivation by adding students' own words to the reliable and oft-used FIT-Choice survey measure. These responses were coded into two schemes, including one that was derived from the FIT-Choice survey measure. The two FIT-Choice responses were compared and found to be fairly similar in the way that the participants' answers are ordered by the two systems. Additional work will help determine the applicability for this open-ended question. However, considering how informative the analyses were with the two coding systems, the FIT-Choice open-ended measure could yield important results in future research studies.
In addition, this examination of preservice teachers' own explanations of why they want to become teachers demonstrates a great deal of positivity and their open-ended answers support their altruistic survey answers. In a time when the teaching profession is being heavily politicized and teachers' ratings of job satisfaction are falling, it is encouraging to hear directly from future teachers that they are choosing the profession to help children and change the world. As we continue our work with preservice teachers, it will be useful to remember what is motivating them and to build upon this positivity in our programs and our classrooms.

Funding
The author received no direct funding for this research.  Want to teach Student indicates wanting to teach, which is more specific than helping children and should also encompass liking children.

Author details
Influenced by previous teacher This category of being influenced by prior teachers encompasses the prior categories, as the influence indicates wanting to teach.  Prior teaching and learning Choice based upon having seen others do the job well.

Ability
The choice to be a teaching is based on perceived ability.
Personal utility value Choice identifies personal benefit.
Intrinsic career value Student is intrinsically motivated and wants to teach.
Social utility value Choice indicates that the student wants to teach and use the future position to benefit society at large. Note. This ranking of categories leaves out the open-ended category of unclassified as the comparison is specifically with the survey categories.