Participation in Ap Computer Science Principles Grows Again

1 INTRODUCTION

Figurer science (CS) holds immense power and influence in our global lodge and economy, and yet from its origins has been and remains an elite and exclusionary field. Indeed, despite recent concerted efforts in the Usa to augment participation in the field, the demographics of the innovators and creators of engineering still do not reflect the racial, cultural, and gender diversity of our society [28]. Exemplifying this imbalance, women represent 57% of the full U.S. workforce but constitute only 26% of the computing workforce [45], and although Black and Hispanic workers constitute 11% and xvi% of the full workforce, respectively, each grouping represents only seven% of the computing workforce [22].

Disparities in computing participation begin early in school settings, as the literacies and experiences shown to motivate youth to enroll in introductory CS courses are often acquired through out-of-school programs predominantly accessed past students from flush families [3, 29]. For people from historically minoritized groups to benefit from participating in and shaping the computing field, and for the computing field to reflect and do good from the rich diversity of the larger society, a broader profile of youth must consider studying CS in higher education, and calculating's distinction as an elite form of knowledge must be dismantled at the elementary and secondary schoolhouse levels [one, 39]. A key focus of such efforts to augment participation in computing (BPC) has been to expand CS didactics opportunities in the United states. In the past 10 years, a series of policy reforms and industry-backed investments take steadily increased access to CS learning opportunities in working-class neighborhoods in the United States through after-school programs [38], summertime coding camps [40], and library makerspaces [9], and through the introduction of computing courses like Exploring Computer science in loftier schools across the country [24, 44].

In this commodity, we examine some other BPC initiative, the relatively new Advanced Placement (AP) form that launched in 2016: Advanced Placement Estimator Science Principles (APCSP). APCSP was designed specifically to increment involvement in CS among a more than diverse population of high school students and to expand the number of students prepared to meet the labor demands of the exponentially growing calculating workforce [18, 68]. We compare 3 groups of students—those who took the new APCSP form, those who took the legacy Advanced Placement Computer Science A (APCSA) class, and students who took both Avant-garde Placement Computer Science (APCS) courses—to examine the extent to which these courses portend undergraduate students' intentions to major in or pursue careers in computing.

two REVIEW OF THE LITERATURE

2.1 The School-Career Computing "Pipeline"

The pipeline metaphor is frequently used to explain demographic disparities in scientific discipline, technology, engineering, and math (STEM) pedagogy and the workforce, as it demonstrates how an essential series of experiences and knowledge starting in childhood are necessary for entry into certain elite professions [43]. The pipeline illustration is particularly pervasive in CS pedagogy research [59], given the trend to view computing education as the primal to calculating majors and jobs [3, 29]. Thus, based on the pipeline metaphor, persistent demographic disparities in the field can be attributed in part to the fact that affluent white male students merely accept more than preparatory CS courses progressively through high schoolhouse and beyond [30, 46].

However, the computing career pipeline metaphor has been critiqued for ignoring the broader systemic forces that perpetually limit opportunities for students to access the coursework and social majuscule necessary to enter and remain in the computing pipeline [44, 59, 66]. Further, the metaphor fails to explain why the gender and racial/ethnic disparities prevalent in computing majors and careers in the U.s.a., Canada, and Western Europe are not observed elsewhere in the world [32]. Accordingly, programmatic efforts to accost diversity in STEM careers using pipeline metaphors take been criticized for forwarding a westernized neoliberal paradigm of education and failing to acknowledge the problematic and undemocratic nature of science itself [42]. Indeed, when it is assumed that at that place is a single pathway to calculating majors and jobs, touchstones like APCS courses tin can human activity every bit gatekeepers that exclude individuals from inbound the pipeline, disproportionately affecting students impacted by a nexus of systemic oppressions like racism, sexism, and classism, who are to the lowest degree likely to have access to such courses [66].

A long-standing offering in the calculating pipeline is APCSA, a loftier schoolhouse elective starting time offered in 1984. With a focus on Coffee programming, APCSA was designed to be comparable to a higher-level introductory class for CS majors and is considered a valuable gateway form for high school students in the Us; it is closely associated with career readiness and interest in pursuing employment in a computing field [iii, 29]. A longitudinal study revealed that 18% of APCSA course-takers eventually majored in a computing-related domain in college, a rate eight times college than AP students of other courses [41]. In add-on, grades earned in AP classes can broaden a course-taker's high school course point average (GPA), and a passing score on the cease-of-course AP exam is often exchanged for undergraduate credit and higher placement in the CS course sequence in college [15].

Despite the office that early CS course-taking plays in shaping subsequent pursuit of CS, scholars have found that computing courses are taught at only about half of the high schools beyond the U.s.a. and that, amidst the classes that are offered, virtually half do not actually innovate students to programming concepts [5]. In fact, although AP courses are available to students across the globe through in-person schools and online programs, APCSA, similar all AP courses, is offered more consistently at high schools in flush neighborhoods in the United States [v, 58, 63], where educators explicitly steer many more students toward computing opportunities [23, 39]. In part because of this inequitable access to courses across schools, 78% of APCSA exam-takers place as white or Asian [68]. To summarize, although students without access to AP classes are already disadvantaged when applying to higher [63], taking and passing APCSA gives other students a competitive edge and accelerates individuals toward related college majors and careers in CS in the United states of america and at the more 400 institutions outside the United States that apply AP courses in their admissions and course-placement procedures [14].

2.2 APCSP to Broaden Participation in Computing

In the aforementioned historic movement of sweeping national BPC efforts in the United States—and with significant back up from the National Science Foundation and unprecedented influence from the engineering science manufacture and industry-backed not-profit partners—the College Board, which administers the AP program, designed a new introductory CS course. APCSP was conceived to address the long-standing racial, gender, and socioeconomic disparities represented in APCSA, in higher educational activity, and in calculating careers in the The states [3]. Higher Lath president and CEO David Coleman articulated that the purpose of APCSP was to "cultivate an involvement in CS among students of all backgrounds" and to prepare class-takers for "the computing jobs that volition help ability future economic growth" [25]. In other words, he explicated before the class launched that two key goals of APCSP were to diversify the group of students who studied calculating in high school (expand access to the pipeline) and increase the number of trained and qualified people who volition eventually work in computing (broaden the pipeline).

To "motivate and brainwash a essentially larger number of students to fulfill the demands of the market while reaching a broader, more diverse segment of the population" [2], APCSP course content emphasizes the big intellectual and fundamental ideas of calculating. The focus on foundational concepts in computing such as creativity, abstraction, information and information, algorithms, programming, the Cyberspace, and the global touch on of computing—rather than APCSA's narrower focus on programming—was designed in part to better concenter and retain students from groups that are underrepresented in calculating, including girls, people of color, and persons with disabilities [18]. Further, different APCSA, which assesses student learning with a summative final AP exam, a singled-out aspect of the APCSP assessment is the inclusion of an applied performance job, centering pick and inventiveness in students' computational work, which is created during classroom instructional time and ultimately accounts for twoscore% of students' final AP score.

2.2.one The Launch of APCSP.

In autumn 2016, APCSP had the largest launch of whatsoever AP course in the College Lath's history [15]. In its inaugural year, 43,780 students from two,500 schools took the APCSP exam [16]; by 2018, this figure had more than doubled to 94,360. Headlines from organizations that promoted diversity initiatives in calculating touted victory, stating: "Girls and minorities intermission records in computer science as fastest-growing groups . . . thanks to the record-breaking launch of the new Information science Principles" [12]. To exist certain, recent information suggests that APCSP attracts a far more than various group of students in terms of gender and race/ethnicity than does the traditional APCSA course [55, 68], although participation amid girls and Black and Hispanic students still does not mirror the diversity of the nation's high school student trunk.

Although there is show that APCSP contributes to both expanding and diversifying the pool of high schoolhouse students learning higher-level CS, at that place is more to explore about what role APCSP plays in ultimately diversifying computing spaces in higher didactics and the workforce. Evidence from the Higher Board [68] suggests that the introduction of APCSP is correlated with a growing number of students majoring in CS in higher, revealing that students who accept APCSP are more likely to major in CS in higher than similar students who graduated high school just prior to the introduction of APCSP (16.9% in 2019 vs. 5.2% in 2016). Their assay matched students graduating high schoolhouse in 2019 with those graduating in 2016 in terms of gender, race/ethnicity, parental pedagogy, high school GPA, and SAT; however, it did not account for additional variables that may correlate with enrollment in APCS courses or otherwise predict major choice, which is a goal of the nowadays study.

3 OBJECTIVES

This empirical enquiry builds on prior studies of higher students who had participated in one or both APCS courses [26, 57, 68] to examine associations between taking APCSA, APCSP, or both APCS courses and students' aspirations to pursue majors or careers in computing. The report includes a set of control variables similar to that used in the College Board [68] written report (east.g., gender, race/ethnicity, parental education, high school GPA, and standardized test scores), in add-on to a wide range of other covariates associated with major and career decision-making. Specifically, using 2017 data from The Freshman Survey (TFS), a nationwide survey of starting time-year higher students administered the year following the formal launch of APCSP, we examined the following research questions:

(i)

How does taking APCSA, APCSP, or both APCS courses stand for with students' intent to major in a computing field or aspire to a career in calculating, and how does this vary by students' gender or race/ethnicity?

(2)

Controlling for proxies for academic achievement (e.g., loftier school GPA, SAT/Deed scores), to what extent does APCS course-taking predict students' intent to major in a computing field or aspire to a career in computing? (a) Are at that place differences in the predictive power of taking only APCSP, but APCSA, or both APCS courses? (b) Does the predictive power of taking each course differ by students' gender or race/ethnicity?

iv METHODS

4.one Data Source and Sample

Information for this study were provided past UCLA'due south Higher Educational activity Research Institute (HERI). Each year, HERI administers TFS to first-year higher students attending institutions that participate in the Cooperative Institutional Research Program. Administered annually since 1966, this comprehensive survey asks respondents nearly their demographic characteristics, values and behavior, bookish and social involvement during their concluding year of high schoolhouse, and their expectations and goals for their higher experience and career. This study relied on information from the 2017 TFS, every bit the APCSP course officially launched in fall 2016 (with a more limited set of schools piloting the course and exam in the 2 years prior) [19]. Thus, these TFS respondents from 168 iv-twelvemonth colleges and universities [64]^one are the beginning course of students who would have had the opportunity to take the fully canonical and more widely offered APCSP class. Among the more than 129,000 respondents in our sample are 6,098 students who took only APCSA, 1,851 who took only APCSP, and 896 who took both APCS courses. Demographically, 58.6% of respondents from the full TFS sample indicated they were women, and the racial/ethnic distribution of the sample was equally follows: 0.two% were American Indian, 14.five% were Asian or Asian American, viii.7% were Black or African American, 8.7% were Hispanic or Latinx, 55.3% were white, 11.9% indicated two or more races/ethnicities, and 0.viii% were another race/ethnicity.

4.two Social Cognitive Career Theory

We employed social cerebral career theory (SCCT) [34] every bit a framework to guide our investigation of the relationships between APCS grade-taking and intentions to pursue a computing major or career. Based on Bandura's social cerebral theory [4], SCCT models how an individual's personal characteristics (e.m., race/ethnicity, gender, academic achievement) and their learning experiences might interact with their self-efficacy beliefs, expectations, interests, and goals, to ultimately shape their bookish and career aspirations [34–36]. The model accounts for a wide variety of pupil identities and conditions that affect students' career evolution and has been found to be helpful for framing educational and career development in STEM fields broadly [11, 62, 67] and in specific fields like computing [33, 57] and engineering [27, 31]. Additionally, scholars have used SCCT to explore how secondary and postsecondary schools might increase the participation of students historically marginalized in STEM education [21, 47, 53, 65]. Our written report builds on this prior research by using SCCT to investigate the potential of APCS course-taking in increasing the racial/indigenous and gender diversity in computing and technology, which is a stated goal of the APCSP grade [19]. Come across Figure 1 for our accommodation of the model.

Fig. 1.

Fig. 1. SCCT model. Adapted from Lent et al. [34].

We practical SCCT to control for a broad variety of experiences, behavior, and interests that inform two different option goals, divers every bit the "intentions, plans, or aspirations to appoint in a detail career direction" [34] among undergraduate students: (1) their intent to major in computing and (2) their aspirations for a career in computing/technology. However, for the purposes of this study, we were most interested in the role of item learning experiences: having taken APCSA, APCSP, or both APCS courses. The SCCT framework suggests that person inputs (e.g., race/ethnicity, gender) and background contextual affordances (e.g., family income, parent education) often dictate students' access to and perceptions of learning experiences, and it is such learning experiences (like APCS courses) that inform individuals' cocky-efficacy, or beliefs about how well one will execute tasks or be successful in a specific domain [34]. Thus, the model suggests that the greater an private's self-efficacy in a task like estimator programming, for example, the more likely they will be to have more positive result expectations of majoring or working in a computing field successfully. Self-efficacy and upshot expectations also influence students' interests or enjoyment in participating in activities, especially those related to the choice goal in question [34]. Consequently, these interests inform students' choice goals. Finally, the framework accounts for the fact that students develop interests, goals, and aspirations near their careers among detail contextual influences.

4.three Measures

Our logistic regression analyses (described in the following) relied on two dependent variables: (1) students' intent to major in calculating and (ii) students' intent to work in a computing or engineering profession. The computing major variable was created by aggregating 90 options for academic major into a dichotomous measure for intended computing major. Students were considered computing majors if they indicated their probable major was "information science," "estimator/direction information systems," or "other math/computer science." If they listed whatsoever other major, they were considered a non-calculating major. Similarly, the dependent variable measuring career involvement in computing and engineering aggregated the 64 career options on the survey into a dichotomous variable that represented students' interest either in a career in computing and technology (which included "computer programmer/developer," "reckoner systems annotator," or "web designer") or an interest in any other career.

Independent variables for both regressions were selected and blocked based on SCCT [34–36] (run across Appendix Table A.ane). The person inputs block included both personal traits and values, and included self-reported social identities (gender^{2 ,} race/ethnicity^three, first-generation status), too as students' belief that programming is important for most careers. The block representing groundwork contextual affordances controlled for boosted factors that might influence students' goals and beliefs [34], and included measures of family income, parent education, and whether or non students had at to the lowest degree ane parent in a Stalk profession. The learning experiences block consisted of measures of academic achievement in high school, such as GPA, SAT/ACT score(s), and time spent writing computer code. This block likewise included our key independent variables of interest: the three dichotomous measures indicating whether students self-reported having taken APCSA, APCSP, or both APCS courses.

The self-efficacy block included measures of students' confidence in their computer programming skills, as well equally constructs measuring their academic self-concept, science self-efficacy, and science identity, among others. The result expectations block captured students' beliefs well-nigh what might happen should they make sure major or career choices, and information technology was the only block that differed between the models. In the regression predicting students' likely major, we used students' confidence that they would remain in their current major of interest, whereas the model examining students' career aspirations included a measure out representing students' confidence that they would maintain their current career plans.

Variables in the interests block included measures representing students' reasons for attending college, such as to make more than money or to gain an appreciation of ideas. Nosotros likewise used a construct representing students' social agency, or the degree to which students are invested in the social welfare of others, as well as measures indicating students' involvement in raising a family, identification every bit artists, or as goal-driven strivers. The contextual influences block included institutional characteristics, such as a measure of institutional selectivity and whether the college is public or individual. Finally, nosotros likewise controlled for students' educational degree aspirations. In sum, SCCT guided our selection of contained variables and enabled usa to examine the predictive power of APCS courses in light of variables understood to predict major and career choice, thus distinguishing this study from related research conducted by the Higher Lath [68].

4.3.1 Constructs and Gene Analysis.

To reduce the number of observed variables and to capture unobserved latent constructs, we performed exploratory factor assay using primary axis factoring with Promax rotation. These analyses were based on factors used in prior research on undergraduate calculating major pick [56] and the impact of college on students [54]. All factors met the reliability thresholds of a Cronbach's alpha of .65 or higher, and items had individual factor loadings greater than .4. We also made use of existing CIRP constructs developed and validated by researchers at the College Pedagogy Research Constitute [17]. These constructs were developed using item response theory, and thus we do not report Cronbach'south alpha or factor loadings for these items [60]. Farther descriptions of factors and constructs and the individual items therein can exist establish in Appendix Table A.1.

4.3.ii Missing Data.

Nosotros used the multiple imputation procedures in SPSS 25 to address missing data. Multiple imputation is currently the almost robust method when data is not missing at random [10]. The procedure involved making a predetermined number of imputations or estimates for each missing value based on existing values in the dataset, then pooling all of the estimates to business relationship for bias in the estimates. Demographic characteristics and our dependent variables were not imputed but were included in imputation models to aid in predicting values of other missing variables. Results reported in Tables 1 and 2 are based on the pooled estimates across 50 imputations. Although there is currently no way to pool estimates of commonly reported fit indices for logistic regression (east.g., pseudo R-squared, Hosmer-Lemeshow), indices for all 50 imputations are available upon request.

Table one. Logistic Regression Predicting Intent to Major in CS Versus All Other Majors (n = 129,241)

• b=regression coefficient in log-odds; SE=standard error of the coefficient; Exp(B)=odds ratios.

• ^aLatent construct; meet Appendix Table A.1.

• *p < .01. **p < .001.

Table 2. Logistic Regression Predicting Career Aspirations in Calculating and Engineering science Versus All Other Careers (due north = 129,241)

• b=regression coefficient in log-odds; SE=standard mistake of the coefficient; Exp(B)=odds ratios.

• ^aLatent construct; come across Appendix Table A.1.

• *p < .01. **p < .001.

4.4 Descriptive Assay

To address our first enquiry question, cross tabulations were run to compare the pct of students who indicated interest in pursuing calculating majors or careers every bit it varied past whether they took APCSA only, APCSP merely, both APCS courses, or neither APCS grade. This analysis was conducted separately past gender and race/ethnicity (Figures 2 and iii).

Fig. 2.

Fig. ii. Per centum of first-twelvemonth undergraduate sample planning to major in calculating, by gender, race, and AP course-taking, fall 2017.

Fig. 3.

Fig. 3. Percentage of beginning-year undergraduate sample planning to pursue a career in calculating, by gender, race, and AP course-taking, fall 2017.

iv.five Regression Assay

To address our 2nd inquiry question and its subparts, we ran one logistic regression to estimate the relationship between APCS course-taking and showtime-year college students' interest in majoring in calculating (see Table 1) and another logistic regression to explore students' career aspirations (see Table 2). For each regression, we first ran a main furnishings model (Model ane) that controlled for a diverseness of possible determinants and influences on the dependent variables. We so ran a model that included the main effects and 18 interaction terms (i.due east., each of the five race/ethnicity variables by the 3 AP class variables; gender by the 3 AP class variables) (Model ii).

5 RESULTS

5.i Inquiry Question 1: Descriptive Results

Our offset research question focused on how taking APCSA, APCSP, or both APCS courses corresponds with entering college students' interest in pursuing computing majors and careers beyond categories of gender and race/ethnicity. As shown in Figure 2, interest in pursuing a computing major among all students was highest among those who took both APCS courses (37.5%), followed past students who took APCSA simply (28.ii%), those who took APCSP only (16.7%), and those who took neither APCS course (3.one%). This full general blueprint held across gender and racial/indigenous groups.

A similar pattern emerged with respect to involvement in a computing career (come across Figure 3), as interest among all students was highest among those who took both APCS courses (31.2%), followed by those who took APCSA only (23.2%), those who took APCSP only (xiv.5%), and those who took neither APCS course (2.5%). Once again, this pattern generally held across gender and racial/ethnic identities.

5.two Inquiry Question 2: Regression Results

The results presented thus far reveal, non surprisingly, a higher degree of involvement in computing majors and careers among students who took APCS courses, with fifty-fifty more than interest amid those who took APCSA simply or both APCS courses. An of import question is whether these associations concur in one case nosotros accounted for the full range of independent variables included in this study. As such, we now plow to research question 2.

In presenting these results, we focus primarily on how the 3 APCS variables predicted major and career aspirations, but because the role of APCS course-taking must be understood in the context of which other variables were meaning, nosotros begin with an overview of such variables. Farther, because we found a high degree of overlap in the variables significant in both the major and career regressions, the following section summarizes key findings across both models. In Table 1 (major) and Table 2 (career), pregnant main effects are shown in Model i, whereas interaction furnishings are presented in Model 2. Given our large sample size, nosotros merely discuss variables significant at p < .01.

5.ii.one Significant Findings Across Major and Career Aspiration Regressions.

Across both models, the likelihood of planning to pursue a calculating major or career was greater among students who identified as men, Asian, Black, Hispanic, or multi-racial. Coming from a lower-income family or having at least ane parent with a Stem career also predicted an increased interest in computing majors and careers. Computing aspirations were significantly lower among students who intended to earn graduate/professional degrees, reported college social cocky-concept, or who held stronger values and orientations effectually social agency, the arts, or family unit. Goals and values that were more predictive of aspirations for computing majors and careers included attending higher with the goal of making more money or to become grooming for a specific career. Non surprisingly, many of these significant variables aligned with findings in prior research exploring CS major aspirations using similar TFS data [56].

Most of the additional variables that were significant across both regressions represented aspects of learning experiences and self-efficacy that were not bachelor in the prior study using these data [56] only chronicle specifically to the calculating and Stem domain. For instance, reporting higher cocky-ratings of computing skills or having written computer code frequently in the past year predicted greater interest in computing majors and careers. More broadly, results revealed greater involvement in calculating majors and careers among students who reported higher levels of science identity but lower levels of science self-efficacy. In addition, greater numbers of AP math and AP scientific discipline courses taken—other than APCS—negatively predicted involvement in computing majors and careers, suggesting that students who have a broader range of advanced Stem courses in high school are less inclined to pursue computing majors and careers.

We at present plow to the main focus of the present study, which is the extent to which taking APCS courses predicts offset-year college students' major or probable career, above and beyond the effects associated with other variables in the model. Farther, we examined how gender or race/ethnicity moderated the relationships betwixt taking ane or both APCS courses and students' interest in computing majors and careers.

v.2.2 APCS Class-Taking Predicting Computing Major.

The logistic regression predicting intent to major in calculating (see Table 1) revealed differences in the salience of our three APCS variables (taking just APCSA, only APCSP, or both APCS courses). Every bit shown in the main effects model (Model 1), students who took only APCSA were 38% more likely to signal CS as their likely major (odds ratio: 1.379) than those who took no APCS courses, even when decision-making for other relevant blocks of variables (Person Inputs, Groundwork Contextual Affordances, Learning Experiences, etc.). Similarly, students who took both APCSA and APCSP were 89% more than likely (odds ratio: 1.893) than students who took no APCS courses to indicate intent to major in CS. However, taking only APCSP was non a pregnant predictor of intent to major in CS.

five.2.two.i Major by Gender Interactions.

Later running the master effects model, we tested whether relationships betwixt taking APCS courses and majoring in computing were moderated by students' gender, seeking here to understand if the predictive power of APCS courses might differ between women and men. Equally shown in Model two of Table 1, all three interactions were significant. In other words, taking APCSA, APCSP, or both APCS courses were all more positively associated with women's aspirations to major in computing than they were for men. Notably, given that taking only APCSP was not meaning in the main effects model, the interaction with gender suggests that taking the class was uniquely associated with women's intent to major in computing.

v.ii.2.2 Major past Race/Ethnicity Interactions.

Every bit shown in Model 2 of Table 1, there were 2 significant interactions between APCS course-taking and students' racial and ethnic identities. Although taking only APCSA was a positive predictor of students' intentions to pursue a CS major for all students generally, the significant interaction term indicated that APCSA was a more than salient predictor of majoring in computing for white students than for Asian students. We besides found that taking both APCS courses was a stronger positive predictor of pursuing a major in CS for white students than for Asian students.

5.2.3 APCS Grade-Taking Predicting Computing or Applied science Career Aspirations.

Using the aforementioned independent variables as in the preceding assay of major aspirations, we ran 2 models to predict students' aspirations to piece of work in computing or technology fields versus all other career options (see Tabular array two). Similar to the findings predicting students' intent to major in computing, Model 1 shows that taking only APCSA or both APCS courses positively predicted students' interest in a computing career (although less then than these courses predicted students' interest in a computing major). Every bit shown in Table 2, students who took but APCSA were 27% more probable to indicate a probable career in computing (odds ratio: ane.269) than students who did not take any APCS courses. Students who took both APCSA and APCSP were 68% more likely (odds ratio: 1.682) than students who did non take whatsoever APCS courses to indicate aspirations to work in computing and technology. Notably, taking only APCSP did not significantly predict students' interest in pursuing a computing career.

5.2.3.one Career by Gender Interactions.

We also explored whether the clan between APCS class-taking and students' career aspirations was moderated past students' gender. We ran 3 cross-product interactions between each APCS form-taking variable and the gender variable. As shown in Model 2 of Tabular array 2, all three interactions were significant. In other words, taking APCSA, APCSP, and both APCS courses appeared to have a stronger positive association with women's calculating career aspirations than with those of men. Although the principal effect of taking only APCSP was not significant in predicting computing career aspirations, the meaning interaction term suggests that APCSP may play a office in encouraging computing career interests for women in detail.

5.two.3.2 Career past Race/Ethnicity Interactions.

We also used interaction terms to explore whether the predictive power of taking any of the APCS course offerings on calculating career aspirations was chastened by students' race or ethnicity. Equally seen in Tabular array 2, simply ane of the interaction terms emerged as significant in predicting a student's computing career aspirations. Specifically, the positive association between taking simply APCSA and students' aspirations to work in computing and tech was stronger amid white students than among Asian students.

6 DISCUSSION

Interpretations of these findings depend on how ane perceives the purpose of APCS courses. If these courses are viewed strictly as the chief entry point in the calculating pipeline, scholars have already found that the addition of APCSP appears to have contributed to broadening participation by increasing the number of students who are taking at least one APCS course [13]. Withal, our study examined first-twelvemonth higher students' longer-term commitment to computing past assessing whether APCS course-taking was associated with aspirations to pursue degrees or careers in calculating fields. After controlling for the predictive ability of about iii dozen variables reflecting unlike facets of SCCT, the results revealed that taking the traditional APCSA course, whether alone or in addition to APCSP, was associated with an increased likelihood of planning to major in or pursue careers in computing. Notwithstanding, taking only APCSP—although positively correlated with computing major and career option based on the descriptive analysis—did non remain a significant predictor of these outcomes in one case the total range of independent variables was controlled.

Our findings did, nevertheless, advise gender and racial/ethnic variations in the predictive power of APCS courses on students' computing majors and career intentions. For instance, taking any of these courses (APCSA, APCSP, or both APCS courses) was more strongly associated with women's plans to pursue computing majors and careers than was the case for men. The finding regarding APCSP is especially noteworthy, as taking only APCSP was non predictive of computing intentions for our total sample. Thus, it is possible that the experience of taking APCS courses—including APCSP—uniquely encourages high school girls to consider computing pathways. Although causality cannot be determined from these data, it is worth noting that the College Board [68] reported a similar finding with respect to gender, albeit with a more than limited set of control variables. Further, our race/ethnicity interactions revealed that taking APCSA (whether lonely or in addition to APCSP) related more strongly to longer-term computing intentions for white students that for Asian students. Taking APCSP on its own, however, did not yield any differences in predictive power by students' race/ethnicity. In other words, across all racial/ethnic groups, computing major and career aspirations appeared unrelated to APCSP grade-taking when the full range of variables were controlled.

vii LIMITATIONS

Information technology is important to acknowledge several limitations to this written report. Get-go, our sample was restricted to offset-time, full-time undergraduate students attending the 168 four-year colleges and universities that participated in the 2017 TFS. Therefore, findings may not reflect the feel of APCS course-takers who attended community colleges or other non-participating institutions, who were enrolled office-time, or who did not enter college in autumn 2017. Our analyses would take benefited from a more representative sample of institutions that included ii-year colleges, as community colleges can be a critical site of grooming for students who somewhen earn degrees in calculating [8, 37]. Further, only a small number (ix%) of U.Southward. public or individual loftier schools even offered APCSP in the 2016–2017 school year, narrowing the sample of students [sixteen].

Second, our data were cantankerous sectional, such that information on students' prior APCS course-taking was collected at the same time every bit their major and career intentions were measured. Therefore, we cannot say whether or to what extent the APCS courses led to students' interests in calculating majors or careers. Indeed, information technology is probable that prior interest in calculating led certain students to enroll in APCS. As such, despite the number of control variables included in this written report, we could non account for self-option bias and causality could not be determined. In addition, the cross-exclusive nature of the data means that career and major aspirations were measured simply at the signal of college entry, despite the reality that students' involvement in pursuing computing fields may alter significantly during college [33, 55]. Relatedly, this report did non account for the fact that many students who pursue non-computing majors and careers really incorporate computing concepts in their called fields (bioinformatics, geoinformatics, etc.); this level of nuance could not be gleaned from the information used in this report.

Third, the sample only reflected students who had taken APCS courses prior to entering higher in fall 2017. This time frame represents only 1 twelvemonth subsequently the official roll-out of APCSP in 2016–2017, although some students may take taken APCSP even earlier during the 2-yr pilot phase. Thus, this study reflects the early stages of the APCSP course and does not account for changes that may accept occurred in course content or delivery as more than schools have adopted the course, as enrollments have continued to grow, and every bit instructors may have evolved their pedagogical approaches. We also practice non know from these information the specific year in which students enrolled in APCS courses, nor which form they took first if they took both APCSA and APCSP.

A terminal gear up of limitations relates to the variables that were or were not included in our study. First, non all elements of SCCT are represented by the variables nosotros included. Second, although SCCT has been widely used as a career development framework, we might also question whether it adequately speaks to the experiences and pathways of all students, especially those who do non follow prescriptive and normative school-to-career pathways. Including other variables into our model or using a different framework might have yielded contrasting results. Finally, our utilize of a relatively large number of command variables suggests that our regression results may exist a conservative estimate of the association between APCSP grade-taking and calculating major and career intentions. In particular, we doubtable that the distinction between our written report's findings and that of the College Board [68] (as it relates to the predictive ability of APCSP) results from our inclusion of a larger range of covariates, including students' computing experiences and self-efficacy.

eight DIRECTIONS FOR FUTURE Inquiry

The limitations described previously point the manner to important new directions for hereafter inquiry on the office of APCS courses and how students' interests in computing majors and careers develop. First, it is important to engage in research that deepens our understanding of what leads high school teachers and advisors to recommend that students take 1 APCS course over another, and specifically to learn whether the current two-grade APCS structure might, peradventure unintentionally, crusade some students to exist "tracked" to a particular APCS course based on assumptions of ability or course rigor. If APCSA portends longer-term retention in the computing pipeline, more so than APCSP, analyses of who takes which APCS course when both are offered (and which course they take first) would help us consider the extent to which inequities are being reproduced in exercise. This is especially important in light of the fact that taking APCSP positively correlates with computing aspirations among women, merely it does non appear to boost computing major or career interests among students of color in particular.

With respect to students taking both APCS courses, this report showed that they had a much higher likelihood of intending to pursue computing majors and careers than students taking either APCSA or APCSP solitary. The contempo College Lath study [68] conveys a similar finding. We believe that a deeper understanding of students who take both courses would add together pregnant value to BPC efforts. Although both courses are considered introductory computing classes—and some assume that APCSP serves every bit a precursor to the more programming-focused APCSA course—it is possible for students to enroll in the two classes meantime, something that ambitious college-going aspirants with access to both courses might elect to do, for the same benefits of augmenting high school transcripts with AP courses.

Next, future inquiry should examine how many and which groups of students take APCS courses in various settings and formats. Our preliminary analyses of TFS respondents' high school IDs showed that some students who reported having taken APCSP attended schools that did not offer the form, at least non as indicated by the College Board.^four Students may have taken APCSP online, at other schools, or as an independent report. Ideally, futurity research on APCS would account for variations in course format while besides considering the characteristics of schools offering the form, besides every bit information on course instructors and their pedagogical approaches. Adding such dash to a study similar the present one could shift, or at least meliorate contextualize, our conclusions near APCSP grade-takers' career and major aspirations.

Chiefly, scholars aimed at determining causal connections between APCS courses and student outcomes would benefit from information that better account for myriad reasons that students enroll in these courses. Although our models controlled for students' motivations for attending college and for their major/career choices, the TFS data did not include students' own reasons for enrolling in APCS courses. Possible reasons include a desire to proceeds valuable skills in their ultimate quest for computing degrees, develop an understanding of basic computing concepts, and/or learn nearly the broader office of calculating in club. Students' motivations for enrolling may also be highly dependent upon communication of teachers, counselors, family members, or peers. Further examination into students' reasons for taking APCS courses, how these vary amidst students taking APCSA, APCSP, or both courses, and how they differ by students' gender, race/ethnicity, and other traits such as family income, high school type (private vs. public), and academic achievement (e.yard., high school GPA, number of AP courses taken) would enable researchers to meliorate assess the effectiveness of these courses equally part of broader diversification efforts in calculating. Additionally, prior inquiry [50, 51, 61] has also suggested that once enrolled in an introductory computing course, students' motivations and aspirations to major or work in a calculating field may change in response to grade content and pedagogy. A better understanding of how and why these aspirations change and for whom may shed calorie-free on pedagogical approaches, activities, and content that tin can help maintain pupil interest.

Finally, research on the role of APCS courses would also ideally follow students' pathways beyond the point of college entry and consider students' use of computing concepts in their not-calculating majors and careers. In doing and then, researchers might also consider outcomes across those frequently associated with the pipeline metaphor (i.e., major or career intentions), besides equally include others, such every bit the want to develop an appreciation of technology in society or develop confidence or self-efficacy in programming. Researchers might as well think fifty-fifty farther out of the box past examining the potential of APCS courses to encourage a sense of agency and empowerment among students, specially for historically minoritized populations [52].

9 Conclusion

APCSP represents an important effort past educators, the College Lath, and technology manufacture leaders to broaden the multifariousness of loftier school students who understand foundational computing concepts, appreciate the role of applied science in order, and ultimately may seek further educational activity and careers in computing fields. Its introduction several years agone marked a critical acquittance that to diversify computing in the U.s., CS instruction needs to extend across the primarily coding-based emphasis of the traditional APCSA form. This study sought to understand the extent to which having taken one or both these APCS courses is associated with inbound college students' involvement in computing majors and careers. Our findings suggested a significant correlation betwixt APCSA and longer-term calculating aspirations, merely that taking APCSP on its ain was not predictive of students' intentions to major or seek employment in computing, at least when we deemed for the full range of control variables. Yet, we did discover that taking APCS courses was significantly more predictive of women's interests in computing-related majors and careers than those of men, and that APCSP did predict women's longer-term calculating interests despite the fact that it was not meaning for the full sample nor for students of color in particular.

Although the findings raise a number of questions for educators and the Higher Board regarding who takes which course and why, consideration of these courses ought non exist in a vacuum. Indeed, discussions on the part of APCS courses must consider that both systems of pedagogy and CS equally a field reside inside and replicate broader systems of oppression such as sexism, racism, and other forms of exclusion [six, 20, 49, 52] and that technological innovations made by computer scientists have further amplified racial hierarchies by hiding, accelerating, and deepening discrimination nether a guise of neutrality [7, 48]. As such, APCSP is not—and should not be expected to exist—a panacea for the structural oppressions that persist in schooling and likely in calculating courses. Further, the true value of APCSP and other diverseness and inclusion efforts cannot be measured solely within the school-career computing pipeline paradigm. Thus, rather than because APCS courses within a fixed and narrow vocational pathway, it is important to explore how the pipeline itself serves to exclude students, particularly students of colour. Expansive efforts to re-envision and transform computing education environments in the Us and elsewhere where broadening participation in calculating is a business concern must attend to how all students can be better welcomed into and supported in calculating courses.

A APPENDIX

1. ¹ Detailed demographic characteristics of APCS class-takers amid the 2017 TFS respondents can be found in our before work [57].

   Footnote

2. ² The TFS asks students to identify as either male or female. In this article, nosotros refer to respondents as men or women as an indication of gender identity rather than biological sexual practice, despite the fact that this variable likely does not capture all students who place equally men or women or any gender identities across this binary.

   Footnote

3. ^three The TFS aggregates students' self-selected racial identities into seven broad categories (shown in Appendix A). We recognize that these groupings may reveal racial/ethnic trends in our data while simultaneously obfuscating whatever heterogeneity that exists within each group and among individuals who exercise not place with whatsoever of the racial/ethnic groups named in the survey.

   Footnote

4. Due to sample size restrictions, only four racial/ethnic groups are reported in the descriptive analyses: Asian, Black, Hispanic, and white.

5. ⁴ This observation is based on supplemental analyses we conducted past matching the survey respondents' high schoolhouse IDs with data provided to us past the College Board on which schools offered APCSP between fall 2013 and spring 2017.

   Footnote

REFERENCES

1. [1] Apple tree Michael West. . 1993. What post-modernists forget: Cultural capital and official noesis. Curriculum Studies 1, 3 (1993), 301–316. DOI: DOI: https://doi.org/10.1080/0965975930010301 Google Scholar Cross Ref
2. [two] Arpaci-Dusseau Andrea C., Astrachan Owen 50., Barnett Dwight, and Bauer Matthew . 2013. Computer science principles: Analysis of a proposed advanced placement form. In Proceedings of the 44th ACM Technical Symposium on Informatics Education . ACM, New York, NY, 251–256. DOI: DOI: https://doi.org/x.1145/2445196.2445273 Google Scholar
3. [3] Aspray William . 2016. Race and Computing: A History of Efforts to Broaden Participation of African Americans, Hispanics, and American Indians in Computing in the Usa . Springer, New York, NY. Google Scholar
4. [4] Bandura Albert . 1985. Social Foundations of Thought and Action: A Social Cognitive Theory . Prentice Hall, Englewood Cliffs, NJ. Google Scholar
5. [v] Banilower Eric R., Smith P. Sean, Malzahn Kristen A., Plumley Courtney L., Gordon Evelyn M., and Hayes Meredith L. . 2018. Written report of the 2018 NSSME+ . Horizon Research, Inc., Chapel Hill, NC. Google Scholar
6. [6] Beasley Maya A. . 2012. Opting Out: Losing the Potential of America's Young Blackness Aristocracy . University of Chicago Press, Chicago, IL. Google Scholar
7. [7] Benjamin Ruha . 2020. Race later on technology: Abolitionist tools for the New Jim Lawmaking. Social Forces 98, 4 (2020), 1–three. DOI: DOI: https://doi.org/10.1093/sf/soz162 Google Scholar Cantankerous Ref
8. [8] Blaney Jennifer M. . 2020. Broadening participation in computing: The role of up transfer. In Proceedings of the 51st ACM Technical Symposium on Informatics Pedagogy . ACM, New York, NY, 254–260. DOI: DOI: https://doi.org/ten.1145/3328778.3366807 Google Scholar Digital Library
9. [9] Shush John J. . 2014. Makerspaces: A Practical Guide for Librarians . Rowman & Littlefield, Lanham, Md. Google Scholar
10. [10] Carpenter James and Kenward Michael . 2012. Multiple Imputation and Its Awarding . John Wiley & Sons, Hoboken, NJ. Google Scholar
11. [eleven] Carpi Anthony, Ronan Darcy M., Falconer Heather Chiliad., and Lents Nathan H. . 2017. Cultivating minority scientists: Undergraduate enquiry increases self-efficacy and career ambitions for underrepresented students in STEM. Journal of Inquiry in Scientific discipline Teaching 54, 2 (2017), 169–194. DOI: DOI: https://doi.org/10.1002/tea.21341 Google Scholar
12. [12] Partovi Hadi . 2018. Girls and minorities break records in estimator scientific discipline as fastest-growing groups! Medium . Retrieved February four, 2021 from https://codeorg.medium.com/girls-and-minorities-suspension-records-in-reckoner-science-as-fastest-growing-groups-39d23425810e Google Scholar
13. [xiii] Code.org. 2020. Dig deeper into AP estimator science. Retrieved February 4, 2021 from https://code.org/promote/ap Google Scholar
14. [14] College Lath. n.d. AP Around the World. Retrieved Baronial 26, 2021 from https://apstudents.collegeboard.org/ap-effectually-the-globe Google Scholar
15. [15] College Board. 2018. AP Cardinal: AP Information science Principles. Retrieved Feb iv, 2021 from https://apcentral.collegeboard.org/courses/ap-computer-science-principles?course=ap-calculator-science-principles Google Scholar
16. [16] College Board. 2017. States propel students to success in advanced placement. College Board Newsroom . Retrieved August 24, 2021 from https://newsroom.collegeboard.org/states-propel-students-success-avant-garde-placement Google Scholar
17. [17] Cooperative Institutional Research Program [CIRP]. 2017. CIRP Construct Technical Study: 2016-2017 Appendix Construct Parameters . Higher Educational activity Enquiry Institute at UCLA. Retrieved February 4, 2021 from https://www.heri.ucla.edu/PDFs/constructs/Appendix2017.pdf Google Scholar
18. [18] Jan Cuny. 2015. Transforming k-12 calculating education: An update and a call to action. ACM Inroads vi, three (2015), 54–57. DOI: DOI: https://doi.org/10.1145/2809795 Google Scholar Digital Library
19. [nineteen] January Cuny. 2015. Transforming K-12 computing education: AP® computer science principles. ACM Inroads six, iv (2015), 58–59. DOI: DOI: https://doi.org/ten.1145/2832916 Google Scholar Digital Library
20. [20] Cutler Kim-Mai . 2015. East of Palo Alto's Eden: Race and the formation of Silicon Valley. TechCrunch . Retrieved February iv, 2021 from https://techcrunch.com/2015/01/10/east-of-palo-altos-eden/ Google Scholar
21. [21] Fouad Nadya A. and Santana Mercedes C. . 2017. SCCT and underrepresented populations in Stalk fields: Moving the needle. Journal of Career Assessment 25, ane (2017), 24–39. DOI: DOI: https://doi.org/10.1177/1069072716658324 Google Scholar Cantankerous Ref
22. [22] Funk Cary and Parker Kim . 2018. Women and men in Stalk ofttimes at odds over workplace disinterestedness. Pew Enquiry Center . Retrieved February four, 2021 from https://www.pewsocialtrends.org/2018/01/09/women-and-men-in-stem-frequently-at-odds-over-workplace-disinterestedness/ Google Scholar
23. [23] Goode Joanna . 2008. Increasing diversity in Thousand-12 computer scientific discipline: Strategies from the field. In Proceedings of the 39th ACM Technical Symposium on Computer Science Education . ACM, New York, NY, 362–366. DOI: DOI: https://doi.org/10.1145/1352135.1352259. Google Scholar
24. [24] Goode Joanna and Margolis Jane . 2011. Exploring computer science: A example study of school reform. ACM Transactions on Computing Teaching 11, 2 (2011), 1–16. DOI: DOI: https://doi.org/10.1145/1993069.1993076 Google Scholar Digital Library
25. [25] Guynn Jessica . 2015. Code.org targets high schoolhouse estimator science. U.s.a. Today . Retrieved February four, 2021 from https://world wide web.usatoday.com/story/tech/2015/05/14/codeorg-college-board-computer-scientific discipline-loftier-schoolhouse-variety/27304593/ Google Scholar
26. [26] Howard Keith E. and Havard Douglas D. . 2019. Advanced placement (AP) calculator science principles: Searching for equity in a two-tiered solution to underrepresentation. Journal of Informatics Integration ii, 1 (2019), 1–15. DOI: DOI: https://doi.org/10.26716/jcsi.2019.02.ane.1 Google Scholar
27. [27] Inda Mercedes, Rodríguez Carmen, and Peña José Vicente . 2013. Gender differences in applying social cognitive career theory in engineering students. Journal of Vocational Beliefs 83, 3 (2013), 346–355. DOI: DOI: https://doi.org/10.1177/0894845315599253 Google Scholar Cross Ref
28. [28] John June P. and Carnoy Martin . 2017. Race and Gender Trends in Information science in the Silicon Valley from 1980-2015 . Stanford University Eye for Educational activity Policy Analysis, Palo Alto, CA. Google Scholar
29. [29] Kaczmarczyk Lisa and Dopplick Renee . 2014. Rebooting the Pathway to Success: Preparing Students for Computing Workforce Needs in the United States . ACM Educational Policy Committee, New York, NY. Google Scholar
30. [xxx] Klassen Myungsook, Stockard Russell, and Akbari Ali . 2004. Stimulating information technology educational activity amidst underrepresented minorities. In Proceedings of the fifth Conference on It Education . ACM, New York, NY, 278–278. DOI: DOI: https://doi.org/10.1145/1029533.1029611 Google Scholar Digital Library
31. [31] Lee Hang-Shim, Flores Lisa Y., Navarro Rachel L., and Kanagui-Muñoz Marlen . 2015. A longitudinal exam of social cognitive career theory's bookish persistence model amidst Latino/a and white men and women engineering students. Journal of Vocational Beliefs 88, (2015), 95–103. DOI: DOI: https://doi.org/10.1016/j.jvb.2015.02.003 Google Scholar Cross Ref
32. [32] Lehman Kathleen J., Newhouse Kaitlin N. Southward., and Sax Linda J. . 2020. Global participation in undergraduate computing: A review and calendar for research. In The Wiley Handbook of Gender Disinterestedness in Higher Education , Niemi Nancy S. and Weaver-Hightower Marcus B. (eds.). Wiley, 289–313. DOI: DOI: https://doi.org/ten.1002/9781119257639.ch15 Google Scholar Cross Ref
33. [33] Lehman Kathleen J., Wofford Annie M., Sendowski Michelle, Newhouse Kaitlin Northward. S., and Sax Linda J. . 2020. Better late than never: Exploring students' pathways to computing in after stages of college. In Proceedings of the 51st ACM Technical Symposium on Information science Education . ACM, New York, NY, 1075–1081. DOI: DOI: https://doi.org/10.1145/3328778.3366814 Google Scholar Digital Library
34. [34] Lent Robert W., Brown Steven D., and Hackett Gail . 1994. Toward a unifying social cognitive theory of career and academic involvement, selection, and performance. Journal of Vocational Behavior 45, one (1994), 79–122. DOI: DOI: https://doi.org/x.1006/jvbe.1994.1027 Google Scholar
35. [35] Lent Robert W., Brown Steven D., and Hackett Gail . 2000. Contextual supports and barriers to career choice: A social cognitive analysis. Journal of Counseling Psychology 47, 1 (2000), 36. DOI: DOI: https://doi.org/10.1037/0022-0167.47.ane.36 Google Scholar Cross Ref
36. [36] Lent Robert W., Lopez Antonio G.., Lopez Frederick G., and Sheu Hung-Bin . 2008. Social cognitive career theory and the prediction of interests and choice goals in the computing disciplines. Journal of Vocational Behavior 73, 1 (2008), 52–62. DOI: DOI: https://doi.org/10.1016/j.jvb.2008.01.002 Google Scholar
37. [37] Ann Lyon Louise and Denner Jill . 2017. Customs colleges: A resource for increasing equity and inclusion in computer science education. Communications of the ACM 60, 12 (2017), 24–26. DOI: DOI: https://doi.org/10.1145/3152914 Google Scholar Digital Library
38. [38] Maloney J. H., Peppler K., Kafai Y., Resnick M., and Rusk N. . 2008. Programming past pick: Urban youth learning programming with scratch. ACM SIGCSE Bulletin forty, 1 (2008), 367–371. DOI: DOI: https://doi.org/10.1145/1352135.1352260 Google Scholar Digital Library
39. [39] Margolis Jane, Estrella Rachel, Goode Joanna, Holme Jennifer Jellison, and Nao Kim . 2017. Stuck in the Shallow Terminate: Education, Race, and Computing . MIT Press, Cambridge, MA. Google Scholar Digital Library
40. [40] Martin Alexis and Scott Allison . 2013. Engaging underrepresented students in calculator science: Examining the effectiveness of a 5-week information science course in the SMASH summer academy. In Proceedings of the 2103 International Conference on Frontiers in Education: Computer Scientific discipline and Computer Engineering . 362–366. DOI:https://www.boom.org/wp-content/uploads/2013/08/examining_the_effectiveness_of_a_5-week_computer_science_course_in__smash-2.pdf Google Scholar
41. [41] Mattern Krista D., Shaw Emily J., and Ewing Maureen . 2011. Advanced Placement® Test Participation: Is AP® Exam Participation and Performance Related to Choice of Higher Major? College Board, New York, NY. Google Scholar
42. [42] Mendick Heather, Berge Maria, and Danielsson Anna . 2017. A critique of the Stalk pipeline: Young people'southward identities in Sweden and science instruction policy. British Journal of Educational Studies 65, 4 (2017), 481–497. DOI: DOI: https://doi.org/10.1080/00071005.2017.1300232 Google Scholar Cross Ref
43. [43] Moses Michele S., Howe Kenneth R., and Niesz Tricia . 1999. The pipeline and student perceptions of schooling: Good news and bad news. Educational Policy 13, 4 (1999), 573–591. DOI: DOI: https://doi.org/x.1177/0895904899013004005 Google Scholar Cross Ref
44. [44] National Academies of Sciences, Technology, and Medicine. 2021. Cultivating Interest and Competencies in Computing: Authentic Experiences and Design Factors . National Academies of Sciences, Engineering, and Medicine, Washington, DC. DOI: DOI: https://doi.org/x.17226/25912 Google Scholar
45. [45] National Center for Women and Data Technology. 2019. NCWIT Fact Canvass. Retrieved February 4, 2021 from https://www.ncwit.org/ncwit-fact-sheet Google Scholar
46. [46] National Scientific discipline Foundation. 2014. College board launches new AP computer science principles course. News Release 14-168. Retrieved February 4, 2021 from https://world wide web.nsf.gov/news/news_summ.jsp?cntn_id=133571 Google Scholar
47. [47] Navarro Rachel L., Flores Lisa Y., Lee Hang-Shim, and Gonzalez Rebecca . 2014. Testing a longitudinal social cognitive model of intended persistence with engineering students across gender and race/ethnicity. Periodical of Vocational Behavior 85, 1 (2014), 146–155. DOI: DOI: https://doi.org/ten.1016/j.jvb.2014.05.007 Google Scholar Cross Ref
48. [48] Noble Safiya Umoja . 2018. Algorithms of Oppression: How Search Engines Reinforce Racism . NYU Press, New York, NY. Google Scholar Cross Ref
49. [49] O'Mara Margaret . 2020. The Code: Silicon Valley and the Remaking of America . Penguin Books, London, Britain. Google Scholar
50. [l] Peteranetz Markeya S., Flanigan Abraham Eastward., Shell Duane F., and Soh Leen-Kiat . 2016. Perceived instrumentality and career aspirations in CS1 courses: Alter and relationships with achievement. In Proceedings of the 2016 ACM Conference on International Calculating Education Research (ICER'16) . ACM, New York, NY, thirteen–21. DOI: DOI: https://doi.org/x.1145/2960310.2960320 Google Scholar Digital Library
51. [51] Peteranetz Markeya S., Flanigan Abraham E., Shell Duane F., and Soh Leen-Kiat . 2018. Career aspirations, perceived instrumentality, and achievement in undergraduate calculator scientific discipline courses. Contemporary Educational Psychology 53, (2018), 27–44. DOI: DOI: https://doi.org/10.1016/j.cedpsych.2018.01.006 Google Scholar Cross Ref
52. [52] Ryoo Jean J., Tanksley Tiera, Estrada Cynthia, and Margolis Jane . 2020. Have infinite, make infinite: How students utilise information science to disrupt and resist marginalization in schools. Computer Science Educational activity 30, 3 (2020), 337–361. DOI: DOI: https://doi.org/10.1080/08993408.2020.1805284 Google Scholar Cantankerous Ref
53. [53] Sahin Alpaslan, Ekmekci Adem, and Waxman Hersh C. . 2017. The relationships amidst loftier school STEM learning experiences, expectations, and mathematics and science efficacy and the likelihood of majoring in Stalk in college. International Periodical of Science Education 39, 11 (2017), 1549–1572. DOI: DOI: https://doi.org/10.1080/09500693.2017.1341067 Google Scholar Cross Ref
54. [54] Sax Linda J. . 2008. The Gender Gap in College: Maximizing the Developmental Potential of Women and Men . Jossey-Bass, San Francisco, CA. Google Scholar
55. [55] Sax Linda J., George Kari L., Wofford Annie M., and Sundar Sarayu . 2019. The tech trajectory: Examining the role of higher environments in enhancing a diverse pipeline to computing careers. Newspaper presented at the annual meeting of the Association for the Study of Higher Education (ASHE). Portland, OR. Google Scholar
56. [56] Sax Linda J., Lehman Kathleen J., Jacobs Jerry A., Kanny M. Allison, Lim Gloria, Monje-Paulson Laura, and Zimmerman Hilary B. . 2017. Anatomy of an indelible gender gap: The evolution of women's participation in informatics. Journal of Higher Instruction 88, two (2016), 258–293. DOI: DOI: https://doi.org/10.1080/00221546.2016.1257306 Google Scholar Cross Ref
57. [57] Sax Linda J., Newhouse Kaitlin N. S., Goode Joanna, Skorodinsky Max, Nakajima Tomoko M., and Sendowski Michelle . 2020. Does AP CS principles broaden participation in calculating? An analysis of APCSA and APCSP participants. In Proceedings of the 51st ACM Technical Symposium on Informatics Education . ACM, New York, NY, 542–548. DOI: DOI: https://doi.org/10.1145/3328778.3366826 Google Scholar
58. [58] Scott Allison, Koshy Sonia, Rao Meghana, Hinton Laura, Flapan Julie, Martin Alexis, and McAlear Frieda . 2019. Information science in California's Schools: An Assay of Admission, Enrollment, and Equity . Kapor Centre, Oakland, CA. Google Scholar
59. [59] Scott Kimberly A. and Elliott Steve . 2020. STEM diversity and inclusion efforts for women of color: A critique of the new labor system. International Journal of Gender, Science and Technology xi, 3 (2020), 374–382. DOI:http://genderandset.open.ac.uk/index.php/genderandset/commodity/view/646 Google Scholar
60. [60] Sharkness Jessica, DeAngelo Lisa, and Pryor John . 2010. CIRP Construct Technical Report . Higher Teaching Research Institute. UCLA, Los Angeles, CA. Google Scholar
61. [61] Vanquish Duane F., Soh Leen-Kiat, Flanigan Abraham East., and Peteranetz Markeya S. . 2016. Students' initial class motivation and their achievement and memory in higher CS1 courses. In Proceedings of the 47th ACM Technical Symposium on Computing Science Pedagogy (SIGCSE'sixteen) . ACM, New York, NY, 639–644. DOI: DOI: https://doi.org/10.1145/2839509.2844606 Google Scholar Digital Library
62. [62] Soldner Matthew, Rowan-Kenyon Heather, Inkelas Karen Kurotsuchi, Garvey Jason, and Robbins Claire . 2012. Supporting students' intentions to persist in Stem disciplines: The role of living-learning programs amid other social-cerebral factors. Journal of Higher Education 83, iii (2012), 311–336. DOI: DOI: https://doi.org/10.1080/00221546.2012.11777246 Google Scholar Cross Ref
63. [63] Solorzano Daniel G. and Ornelas Armida . 2002. A critical race analysis of avant-garde placement classes: A case of educational inequality. Periodical of Latinos and Educational activity 1, iv (2002), 215–229. DOI: DOI: https://doi.org/x.1207/S1532771XJLE0104_2 Google Scholar Cross Ref
64. [64] Bara Stolzenberg Ellen, Eagan M. Kevin, Aragon Melissa C., Cesar-Davis Natacha Chiliad., Jacobo Sidronio, Couch Victoria, and Rios-Aguilar Cecelia . 2019. The American Freshman: National Norms Autumn 2017 . College Instruction Research Institute. UCLA, Los Angeles, CA. Google Scholar
65. [65] Szelényi Katalin, Denson Nida, and Inkelas Karen Kurotsuchi . 2013. Women in Stem majors and professional effect expectations: The office of living-learning programs and other higher environments. Enquiry in Higher Educational activity 54, 8 (2013), 851–873. DOI: DOI: ten.1007/s11162-013-9299-2 Google Scholar Cantankerous Ref
66. [66] Vitores Anna and Gil-Juárez Adriana . 2016. The problem with 'women in calculating': A critical examination of the deployment of research on the gender gap in computer science. Journal of Gender Studies 25, 6 (2016), 666–680. DOI: DOI: https://doi.org/ten.1080/09589236.2015.1087309 Google Scholar Cross Ref
67. [67] Wang Xueli . 2013. Why students cull Stem majors: Motivation, high school learning, and postsecondary context of support. American Educational Enquiry Journal 50, 5 (2013), 1081–1121. DOI:https://doi.org/10.3102/0002831213488622 Google Scholar Cross Ref
68. [68] Wyatt Jeff, Feng Jing, and Ewing Maureen . 2020. AP Information science Principles and the STEM and Information science Pipelines . College Board, New York, NY. Retrieved February 4, 2021 from https://apcentral.collegeboard.org/pdf/ap-csp-and-stem-cs-pipelines.pdf?course=ap-calculator-science-principles. Google Scholar

robinsonyouldrals.blogspot.com

Source: https://dl.acm.org/doi/full/10.1145/3479431

Share this post