Prepared for the Real World: New Survey Shows K12 Online High School Better Prepares Students with Career Readiness Skills and Job Success vs. Brick-and-Mortar Public Schools

A recent survey commissioned by Stride, Inc. measured how K12-powered schools prepare students for postsecondary and career success. The survey was administered to young adults who graduated from K12 online schools as well as a nationally representative group of American adults ages 18–29.

Abstract

While several studies have evaluated virtual charter schools in terms of their impact on student achievement, there is scant literature that explores how virtual charters prepare students for later in life outcomes, including college and career readiness. Such questions take on increased importance in the wake of a nascent but growing literature that finds sizable disconnects in school choice programs between achievement and later in life outcomes.

To assess the degree to which virtual schools prepare students for postsecondary and career success, I administer a survey to young adults (ages 18-29) who graduated from virtual charters managed by K12, the largest education management organization in the United States. I also partner with a market research firm to administer the same study to a nationally representative group of American adults ages 18-29. Survey questions assess four constructs related to postsecondary outcomes (college-going culture, postsecondary support, academic behaviors, and communication) and three related to career outcomes (advising, self-concept, and soft skills). Overall, virtual charter graduates report statistically significant advantages when it comes to the three career constructs and two of the postsecondary constructs (academic behaviors and communication), but statistically significant disadvantages when it comes to college-going culture and postsecondary support. Implications and limitations are discussed.

Introduction

Virtual charter schools-charters that exclusively educate students through a digital medium- have existed in the United States for more than two decades. In 2019-20 there were more than 300,000 students enrolled in virtual schools (Molnar et al., 2021). Enrollment increased dramatically the following school year owing to the COVID-19 pandemic and associated school closures, inviting fresh deliberation about what role digital learning and virtual charters should play within the US education landscape (Hannah, 2020; Kamenetz, 2020).

Critics cite comparatively low observed proficiency rates, high mobility, and low graduation rates as reasons to bar the establishment of virtual charters or, in states where they already exist, to tighten regulations or place caps on enrollment. Proponents meanwhile point to the pitfalls of using proficiency and graduation rates as an indicator of school quality, as they tend to provide more information about the type of students that a school serves rather than the quality of the school itself (Chingos, 2013).

Attempts to discern their quality are further obfuscated by suboptimal testing conditions (specifically, narrow testing windows that appear to suppress scores) (Beck, Watson & Maranto, 2019), the degree to which high mobility deleteriously impacts student achievement (Paul & Wolf, 2020) and disproportionate negative selection that appears to negate the validity of more sophisticated evaluation designs, such as student matching (Paul & Greene, 2022).

Challenges of using test scores to evaluate virtual charters aside, recent research offers compelling evidence of a disconnect between how schools of choice effect achievement versus long-run outcomes such as college enrollment and earnings (Hitt, McShane & Wolf, 2018). Of particular note is a quasi-experimental study conducted in Barbados which revealed that families gaining access to their preferred schools is not associated with improved test scores but is associated with improved educational attainment, earnings, and health (Beuermann & Jackson, 2019), leading the authors to caution that “test score impacts may not be the best measure of a school’s impacts on longer-run outcomes.” (p. 12).

Surveys as an Evaluation Tool

Surveys stand out as a potentially important supplemental tool in assessing the quality of virtual schools. Indeed, in surveying families of virtual school students about their satisfaction with their schools, Greene & Paul (2022) note that “Parental satisfaction has stronger predictive power for later life outcomes than do standardized test scores and is less subject to manipulation than are other measures.”

Surveys feature other benefits, including more granular and actionable information that can be provided by quantitative measures of effectiveness (e.g., value-added modeling) and low financial and time cost (Balch, 2015). However, surveys feature some notable limitations. Some research lends credence to the notion that student surveys can be valid and reliable (Wilkerston et al., 2000; Ferguson, 2012) but other studies raise concerns. Wladis and Samuels (2016) for example concluded that “online readiness surveys” had no predictive validity regarding outcomes in postsecondary online courses. Critically, most research on the predictive validity of student surveys has occurred in postsecondary settings, and that research which has occurred in primary and secondary settings has probed validity and reliability as it relates to teacher quality, not school quality. Simply put, the reliability and validity of a high school survey to recent alumni is unknown, and findings must be interpreted with caution. Still, to the degree that other measures of school quality are also imperfect, survey data can provide important context often unattainable from other measures and, in conjunction with other measures (e.g., value-added and classroom observations), a more complete picture of how well virtual schools fulfill one of their vital responsibilities.

Methods

I constructed a survey comprised of 30 Likert-scale questions to measure various aspects of college and career readiness. I piloted the questions with a convenience sample of four young adults (ages 20-24) to assess their clarity, which revealed no issues or concerns. Questions are original but draw inspiration from existing surveys, such as the Los Angeles Unified School District College and Career Readiness Survey (LAUSD, n.d.). Career readiness questions also draw inspiration from popular job search websites that poll industry leaders about the types of skills that such leaders seek in job candidates (Indeed Editorial Team, 2020; Anderson, 2020).

Survey items are combined to measure four constructs related to college readiness and four related to career readiness. The scales, their associated questions, and their Cronbach’s alpha scores are displayed in Table One.

Table One: Career and College Readiness Constructs

College Readiness

College-going Culture	α
Teachers at my high school spoke frequently about the benefits of attending college.	.81
Teachers at my high school positively influenced my decision to attend college.	.79
It felt like students at my high school were expected to enroll in college.	.96

Postsecondary Support	α
High school teachers and/or staff were helpful in answering questions that I had about college.	.88
High school teachers and/or staff helped me decide where to enroll in college.	.90
High school teachers and/or staff were helpful during the college application process.	.93

Academic Behaviors	α
My high school prepared me to read hundreds of pages of text per week.	.80
My high school prepared me to write long essays.	.85
My high school prepared me to take good notes during class.	.85
My high school prepared me for taking final exams that cover all material from the course.	.83
My high school prepared me for pacing the workload of large projects.	.85

Communication	α
My high school prepared me well for class discussions.	.89
My high school prepared me well for speaking up when I need clarification on content or instructions.	.91
My high school prepared me well for corresponding with the teacher/professor about how to succeed in their class.	.90

Career Readiness

Career Advising	α
High school teachers and/or administrators were helpful in guiding me toward a career.	.92
High school teachers/administrators taught me to be ambitious in my career goals.	.93
High school teachers/administrators provided good advice about how to succeed professionally.	.92

Career Self-Concept	α
I am excelling at my current job.	.89
I am optimistic about where my career is heading.	.87
My current job is my dream job.	.85
My current job is a good steppingstone to my dream job.	.81

Soft Skills	α
My high school prepared me to be an analytical thinker.	.95
My high school prepared me for complex problem-solving.	.95
My high school taught me to manage my time well.	.95
My high school impressed upon me the importance of a strong work ethic.	.95
My high school helped me cultivate interpersonal skills.	.95
My high school helped me cultivate leadership skills.	.95
My high school helped me to become a better communicator.	.95

Methods

The survey was administered by K12 to young adults who graduated from one of their managed public charter schools. Overall, 63,341 surveys were distributed to alums who graduated between 2012 2021. Of the 63,341 surveys distributed, 1,339 were completed, for a response rate of 2.1%. I also partnered with a market research firm to administer the same survey to a nationally representative sample of 1,700 young adults (ages 18-29) who graduated high school so that responses from K12 graduates have a sensible comparison group.

There are some notable differences between the two groups of young adults. For example, the average age of the non-Stride students is 23.6 whereas the average age of the responding K12 graduates is 22.2. Moreover, among only 30.2% of K12 respondents are male compared to 47.6% in the comparison group. Two-sided t-tests also reveal significant differences between the two groups in how they identify by race/ethnicity, as seen in Table Two. Analysis features multiple regression estimates to reveal the extent to which these differences might contribute to differences in responses to survey questions.

Table Two: Differences Between K12 Graduates and Comparison Group

Category	K12 sample	Comparison Sample	Difference
Age	22.2	23.6	***
Male	30.2%	47.6%	***
African American	6.7%	13.3%	***
Asian	1.8%	5.5%	***
Hispanic	8.4%	11.1%	**
Native American	1.1%	1.1%
White	58.2%	61.4%	*
HH Income <50k	53.8%	53.9%
HH Income <50k – <100k	31.3%	30.8%
HH Income <100k – <150k	9.3%	9.5%
HH Income <150k – <200k	3.7%	3.6%
HH Income <200k	1.9%	2.2%

The low response rate among graduates of virtual schools raises concerns that responses may not be representative of the general population of virtual school graduates. To probe this concern, I assess differences in observable characteristics between those who responded (n=1,339) to the survey and those who did not respond (n=62,006), as seen in Table Three. While such analysis cannot reveal potentially unobservable differences, it does at least offer some insight into the degree to which survey uptake was not random (Goyder, 1987).

Table Two: Differences Between K12 Graduates and Comparison Group

Category	Respondents	Non-Respondents	Difference
Eligibility for free or reduced-price lunch in senior year of high school (%)	46.4	43.5	**
Course completion rate (%)	94.3	92.2	***
African American (%)	7.7	8.5
Asian (%)	2.7	2.3
Hispanic (%)	13.0	11.8
Native American (%)	1.6	1.1	*
White (%)	51.2	48.7	*
Male (%)	29.9	39.2	***

Two-sided t-tests reveal that the two groups profile similarly in terms of racial and socioeconomic characteristics. Course completion rates- the percentage of enrolled courses that students completed between grades 9 and 12- also indicate that they profile similarly in terms of their high school academic achievement. There is however a notable difference by gender: Only 29.9% of survey respondents were male compared to 39.2% of those who did not respond. The higher uptake of surveys among women is well-documented in academic literature (Curtin et al., 2000; Singer et al., 2000; Smith, 2008) but it nevertheless reinforces that the representativeness of responses must be treated with some caution.

Analysis

Likert-scale questions ranged from one to seven (1-strongly disagree, 2-disagree, 3-somewhat disagree, 4-neutral, 5-somewhat agree, 6-agree, 7-strongly agree). In the analysis that follows, the scores represent the average response on the questions associated with each scale.

College Readiness

Postsecondary education is associated with significantly higher career earnings and job satisfaction, better health, more marriages and fewer divorces, and higher levels of societal trust and community engagement (Black & Smith, 2008; Hout, 2012; Oreopolous & Salvanes, 2009; Psacharopoulos & Patrinos, 2018). These benefits are typically observed even when using quasi-experimental methods. Consequently, policymakers and researchers tend to treat college matriculation and persistence as an important lagging indicator of primary and secondary school quality (Barnes & Slate, 2013; Berliner, 2006; Ravitch, 2010).

College Readiness

The college-going culture subscale reveals a comparatively weak college-going culture in virtual charters. Whereas the virtual school group reports an average score of 4.70, the comparison group reports an average score of 5.31. The difference is statistically significant at the 99% confidence level, and it is not sensitive to additional controls.

Category	K12	Comparison group	Difference
Teachers at my high school spoke frequently about the benefits of attending college.	4.85	5.47	***
Teachers at my high school positively influenced my decision to attend college.	4.63	5.11	***
It felt like students at my high school were expected to enroll in college	4.63	5.28	***

Differences in college-going culture-the degree to which “students find encouragement and help from multiple sources to prepare them with knowledge needed for college success” (McKillip et al., 2013, p. 530)- raise questions about whether such mores have structural or ecological origins. That is, does a digital learning environment inhibit the flourishing of a college-going culture, or do virtual charter schools simply teach a student population that is, on average, less predisposed toward college? Several studies (Beck & Maranto, 2014; Beck et al., 2014; Kingsbury, et al., 2022; Maranto et al., 2021; Paul & Greene, 2022) suggest that virtual charters disproportionately serve at-risk students, hinting that student composition might play an important role in the comparatively weak college-going culture of virtual schools, but future research might explore this question further.

Category	I	II	III	IV	V
Virtual School grad	-.61*** (.08)	-.57*** (.09)	-.61***(.08)	-.60***(.08)	-.58***(.09)
Age	–	-.00(.01)	–	–	-.01(.01)
Male	–	–	-.02(.08)	–	-.05(.09)
African American	–	–	–	.21(.17)	.15(.18)
Asian	–	–	–	.27(.23)	.15(.24)
Hispanic	–	–	–	.30*(17)	.15(.18)
Native American	–	–	–	-.21(.44)	-.46(.47)
White	–	–	–	-29**(.13)	-20(.14)
n	1.555	1.471	1.510	1.550	1.427

It should be noted that there is evidence that measures of college-going culture are not as predictive of postsecondary outcomes as are secondary achievement outcomes (Bryan et al., 2018). Whatever the cause of the comparatively weak college-going culture in virtual schools, it remains unclear to what extent that influences matriculation in or persistence through college.

Postsecondary Support

A separate construct measures the degree to which teachers and staff were supportive to students interested in enrolling in college, a measure perhaps often correlated with college-going culture but also appreciably distinct, as culture is an amalgam of diverse school- and family-level inputs whereas postsecondary support specifically addresses student interactions with teachers and staff. Overall, results are similar to the college-going culture construct, revealing that virtual school students report a significant disadvantage in the degree to which teachers and staff support them in the college application process. Multiple regression estimates indicate that the results are not sensitive to additional controls.

Postsecondary Support	K12	Comparison group	Difference
High school teachers and/or staff were helpful in answering questions that I had about college.	4.73	4.89	**
High school teachers and/or staff helped me decide where to enroll in college.	3.28	4.23	***
High school teachers and/or staff were helpful during the college application process.	3.81	4.52	***

Notably, the biggest difference between the two groups is in response to the question that asks about the degree to which teachers/and or staff “helped me decide where to enroll in college.” Compared to brick-and-mortar schools, virtual schools expect intensive cooperation and assistance from a child’s legal guardian and deputize them as de facto tutors and mentors (Black, Ferdig & DiPietro, 2008; Litke, 1998). It is likely that the intensive participation that parents take on in a child’s educational journey means that they supplant the role of teachers in various ways, including perhaps advising on the college application process. To what extent this probable supplantation effects matriculation and persistence through college stands out as a question worthy of further inquiry.

Category	I	II	III	IV	V
Virtual School grad	-.60*** (.08)	-.61*** (.08)	-.60***(.08)	-.64***(.08)	-.64***(.09)
Age	–	-.03***(.01)	–	–	-.02(.01)
Male	–	–	-.03(.08)	–	-.03(.08)
African American	–	–	–	.60**(.25)	.57**(.27)
Asian	–	–	–	.02(.20)	.01(.20)
Hispanic	–	–	–	.11(14)	.03(.15)
Native American	–	–	–	-.20(.41)	-.34(.44)
White	–	–	–	-13(.10)	-16(.10)
n	1.765	1.686	1.717	1.763	1.636

Academic Behaviors

The academic behaviors construct assesses the degree to which high schools properly instill the type of hard skills required for success in college. Traditionally, evaluators and researchers have relied upon metrics such as state assessment outcomes and courses taken in high school (Gaertner & McClarty, 2015). Such measures have been demonstrated to correlate moderately with postsecondary outcomes (Adelman, 1999; Geiser & Santelices, 2007; Kurlaender et al., 2008; Roderick & Nagaoka, 2005). Goodwin and Hein (2016) specifically assert that high school grade point average and college entrance scores predict only 20 to 25 percent of a student’s college achievement. Plus, as Conley (2007) notes, these metrics are imperfect in that they fail to capture whether students who had the aptitude and work ethic to excel in high school increased their work ethic to meet the comparatively greater demands of college. Indeed, college courses routinely “require students to read eight to ten books in the same time that a high school class requires only one or two.” (p. 6). Not surprisingly, “college faculty consistently report that freshman students need to be spending nearly twice the time they indicate spending currently to prepare for class.” (p. 7). Assessing then whether high school prepared students to take up the type of study habits required for postsecondary success stands out as an important predictive metric.

Academic Behaviors	K12	Comparison group	Difference
My high school prepared me to read hundreds of pages of text per week.	4.84	4.18	***
My high school prepared me to write long essays.	35.35	4.86	***
My high school prepared me to take good notes during class.	5.20	4.19	***
My high school prepared me for taking final exams that cover all material from the course.	5.12	4.86	***
My high school prepared me for pacing the workload of large projects.	5.46	5.02	***

Overall, virtual school graduates report clear advantages in the degree to which high school prepared them with the type of academic behaviors required for postsecondary success. The 5.46 score on the question-“My high school prepared me for pacing the workload of large projects” -denoting an average response that falls between somewhat agree and agree-registers as the highest of all questions asked about postsecondary readiness. While the comparatively autonomous and self-paced learning that occurs in virtual schools presents challenges for many students (Ahn & McEachin, 2017), perhaps the practical limitations on teacher oversight also more closely emulate and better prepare students for the postsecondary academic experience.

Category	I	II	III	IV	V
Virtual School grad	-.56*** (.08)	-.59*** (.08)	-.54***(.08)	-.57***(.09)	-.57***(.09)
Age	–	-.00(.01)	–	–	-.00(.01)
Male	–	–	-.12(.08)	–	-.14(.08)
African American	–	–	–	.07(.28)	.07(.28)
Asian	–	–	–	.-32(.19)	.32*(.19)
Hispanic	–	–	–	.-09(.14)	.09(.14)
Native American	–	–	–	-.35(.40)	-.35(.40)
White	–	–	–	-08(.10)	-08(.10)
n	1.629	1.567	1.586	1.524	1.524

Communication

Willingness to communicate academic questions and concerns to peers and professors is predictive of postsecondary academic success (Rubin & Graham, 1988). Indeed, subjective measures of effective communication predict both postsecondary achievement and persistence (Hawken et al., 1991), whereas communication apprehension has been positively linked to persistence in two studies, one of which also finds that it is predictive of achievement (Ericson & Gardner, 1992; McCroskey et al., 1989).

Communication	K12	Comparison group	Difference
My high school prepared me well for class discussions.	5.34	4.85	***
My high school prepared me well for speaking up when I need clarification on content or instructions.	5.43	4.76	***
My high school prepared me well for corresponding with the teacher/professor about how to succeed in their class.	5.56	4.78	***

Responses from K12 graduates and the comparison group reveal significant advantages for the former in the degree to which they feel high school prepared them for communication competence, “functionally effective interaction appropriate to a given relational context.” (Splitzberg, 1983, p. 323). The results are perhaps surprising and defy the conventional wisdom, which holds that a digital learning platform inhibits proper socialization (Ash, 2009). While this topic is scarcely studied, there is indeed some instructive literature to suggest that the conventional wisdom might be wrong. A white paper commissioned by K12 Inc. (now K12) and undertaken by external researchers relays that students in fully online schools rated better than a comparison group on social skills in self-evaluations and parental evaluations. Moreover, students in full time online schools appear to be highly engaged in social activities outside of school. For example, 68% of parents of students in virtual schools reported that their children played sports with other children at least 2 to 3 times per month. Whether a digital learning environment fosters proper socialization or whether diligent parents can appropriately foster it through outside-school activities remains unclear, but whatever the case, it certainly appears plausible that graduates of fully online schools enter college with superior communication skills on average. As the white paper highlights, the difference between perception and reality regarding socialization in virtual schools might mirror homeschooled children. Whereas the conventional wisdom holds that homeschool children are not well socialized compared to peers, a robust academic literature indicates that the opposite appears to be true.

Category	I	II	III	IV	V
Virtual School grad	-.71 (.08)	-.74(.09)	-.59(.08)	-.68(.08)	-.68(.09)
Age	–	-.01(.01)	–	–	-.01(.01)
Male	–	–	-.05(.08)	–	-.06(.08)
African American	–	–	–	.44(.27)	.43(.29)
Asian	–	–	–	.-06(.19)	.11(.19)
Hispanic	–	–	–	.-19(.14)	.13(.14)
Native American	–	–	–	-.58(.39)	-.71*(.42)
White	–	–	–	-00(.09)	-01(.10)
n	1.663	1.592	1.632	1.659	1.552

Career Readiness

Some students have no intention of attending college or may simply have skills or individual circumstances that leave them better served to directly enter the work force upon graduating high school. This might be especially true in virtual charter schools given the disproportionate negative selection that appears to occur in such settings. To that end, for students who don’t enroll in college, the degree to which high school prepares them for labor force participation is a potentially important if often unappreciated indicator of school quality.

Career Advising

In a review of extant literature, Hughes and Karp (2004) find that studies using a variety of different methods tend to show that students benefit from receiving career guidance while in middle school and/or high school. Critically, these benefits have also been observed in studies that specifically consider computer-based guidance systems, highlighting that, to whatever degree and quality of career advising occurs in virtual schools, it would seem to have the potential to meaningfully influence student behavior or attitudes. These benefits may not only manifest as improved human capital stock, but improved social capital (Hooley & Dodd, 2015).

Career Advising	K12	Comparison group	Difference
High school teachers and/or administrators were helpful in guiding me toward a career.	4.19	3.62	***
High school teachers/administrators taught me to be ambitious in my career goals.	4.50	3.92	***
High school teachers/administrators provided good advice about how to succeed professionally.	4.47	3.83	***

While K12 graduates gave more favorable responses than the comparison group, the overall scores among both groups perhaps raise alarm. Indeed, the responses from K12 graduates fall between neutrality to tepid agreement, whereas the responses from the comparison group fall between neutrality and tepid disagreement. Given the reality that not all students are well served by attending college, the responses elevate concerns that schools are induced by accountability regimes to favor test scores over other considerations that might yield greater benefits, particularly to students who don’t attend college (Darling-Hammond, 2007).

Category	I	II	III	IV	V
Virtual School grad	-.60*** (.09)	-.65***(.09)	-.61***(.09)	-.61***(.09)	-.66***(.10)
Age	–	-.00(.01)	–	–	-.00(.01)
Male	–	–	-.05(.09)	–	-.03(.09)
African American	–	–	–	.01(.31)	.03(.35)
Asian	–	–	–	.-06(.25)	.08(.25)
Hispanic	–	–	–	.-25(.17)	.18(.18)
Native American	–	–	–	-.57(.34)	-.59(.42)
White	–	–	–	-.08(.12)	-.09(.13)
n	1.750	1.653	1.683	1.746	1.591

Career Self-Concept

The notion of self-concept-succinctly, beliefs that individuals hold about themselves- has been foundational to social psychology for more than a century, and it continues to feature prominently into psychosocial research and discourse (Gecas, 1982; Swann et al., 2007). It is often used as a lens within education research (Shavelson et al., 1976), matters related to workforce development (Betz, 1994), and the points at which those topics converge (Fenning & May, 2013; Nasir & Lin, 2013). Taskinen et al. (2013) for example examine the importance of science self-concept and various school-level factors in adolescents’ motivation to select an academic science-related career. Note that there is fierce debate about the predictive utility of measures of self-concept vis-à-vis important outcomes downstream (e.g., academic and career outcomes), but Swann et al. (2007) make a compelling argument that much of the skepticism is attributable to the importance of such concepts occasionally being oversold but that, nevertheless, the predictive utility of self-concept is well-established. ¹

Career Self-Concept	K12	Comparison group	Difference
I am excelling at my current job.	5.48	3.25	***
I am optimistic about where my career is heading.	5.19	3.90	***
My current job is my dream job.	3.60	2.56	***
My current job is a good steppingstone to my dream job.	4.37	2.85	***

K12 graduates report starkly different responses to questions regarding the status and trajectory of their career. Whereas K12 graduates on average fall between “somewhat agreeing” and “agreeing” that they are excelling at their current job and feel optimistic about where their career is going, the comparison group responses ranging from tepid disagreement to neutrality. Indeed, differences to the prompt “I am excelling at my current job” register as among the highest on any question on the survey, outpacing the comparison group by more than one full standard deviation. Differences are not sensitive to the inclusion of additional controls. Precisely why responses on these questions divulge so dramatically is difficult to ascertain, though it certainly strengthens recent arguments that virtual schools might provide value not detected by test scores (Greene & Paul, 2022).

Category	I	II	III	IV	V
Virtual School grad	-.1.52*** (.11)	-.1.60***(.11)	-.1.52***(.11)	-.1.52***(.11)	-.1.61***(.12)
Age	–	-.04***(.01)	–	–	-.04***(.01)
Male	–	–	-.10(.10)	–	-.10(.11)
African American	–	–	–	.10(.27)	.17(.30)
Asian	–	–	–	.-24(.36)	.14(.37)
Hispanic	–	–	–	.-07(.20)	.13(.22)
Native American	–	–	–	-.75*(.45)	-.10(.49)
White	–	–	–	-.02(.14)	-.05(.15)
n	956	858	906	952	813

Soft Skills

Heckman & Kautz (2012) find compelling evidence that soft skills – “personality traits, goals, motivations, and preferences that are valued in the labor market” (Heckman & Kautz, 2012, p. 451)- can make an important difference in a child’s life, and that they are an integral component of adolescent development even though they are not systematically measured in the American public education system. Indeed, soft skills “predict success in life…causally produce that success…(and) programs that enhance soft skills have an important place in an effective portfolio of public policies (Heckman & Kautz, 2012, p. 451). Indeed, soft skills may dwarf hard skills in the degree to which they predict labor market outcomes (Meenu & Kumar, 2009).

Soft Skills	K12	Comparison group	Difference
My high school prepared me to be an analytical thinker.	4.61	3.82	***
My high school prepared me for complex problem-solving.	4.68	3.90	***
My high school taught me to manage my time well.	5.15	3.91	***
My high school impressed upon me the importance of a strong work ethic.	4.96	4.11	***
My high school helped me cultivate interpersonal skills.	4.37	4.00	***
My high school helped me cultivate leadership skills.	4.29	3.94	***
My high school helped me to become a better communicator.	4.59	4.05	***

Overall, K12 graduates report that their high schools performed better than the comparison group in the degree to which they instilled soft skills. Of particular note are the final three questions, which ask about interpersonal, leadership, and communication skills. The higher marks from K12 graduates reinforce the aforementioned evidence that the conventional wisdom about socialization in virtual schools might be incorrect, and they provide evidence that the advantage that virtual charters exhibit in terms of socialization manifest in both college and in the labor force. Moreover, the relatively stronger responses might explain the similar wages between K12 and their comparison group (Table Two), an arguably surprising observation given comparatively weak observed achievement in K12-powered schools, disproportionate negative selection, and comparatively disadvantaged economic background of enrolled students (Scafidi 2022, forthcoming).

Conclusion

Given the limitations of surveys as evaluative tools and the small, potentially non-random sampling that occurred for this study, the findings should be treated with some caution. Still, the results suggest that evaluations of virtual charter schools that rely upon test scores may not be telling the full story, and that virtual charters may in fact be doing a good job of preparing students for college and career readiness in ways not detected by test scores. Future research should analyze household income data from a larger sample of virtual school students to observe whether the results captured here are anomalous, as the results captured here raise the distinct possibility that test scores may have limited predictive validity in virtual schools. In the meantime, states may want to adopt student surveys as a component of their school evaluation systems to ensure that school quality is measured holistically.

• Adelman, C. (1999). Answers in the tool box: Academic intensity, attendance patterns, and bachelor’s degree attainment. Washington, DC: U.S. Department of Education.
• Ahn, J., & McEachin, A. (2017). Student enrollment patterns and achievement in Ohio’s online charter schools. Educ. Res. 46, 44–57.
• Anderson, B. (2020). The most in-demand hard and soft skills of 2020. LinkedIn. Retrieved from https://www.linkedin.com/… .
• Ash, K. (2009). Socialization in virtual education. Education Week. Retrieved from https://www.edweek.org/… .
• Balch, R. (2015). Using student surveys at the elementary and secondary levels. In J. Grissom & P. Youngs (Eds.), Improving teacher evaluation systems: Making the most of multiple measures (pp. 77-88).
• Barnes, W. & Slate, J. (2013). College-readiness is not one-size-fits-all. Current Issues in Education, 16(1), 1-13.
• Beck, D., Egalite, A., & Maranto, R. (2014). Why they choose and how it goes: Comparing special education and general education cyber student perceptions. Computers & Education, 76(1), 70-79.
• Beck, D. & Maranto, R. (2014). Empowering teachers? An exploratory study of personnel practices in virtual charter schools in the United States. Journal of Open, Flexible, and Distance Learning, 18(2), 59-82.
• Beck, D., Waston, A. & Maranto, R. (2019). Do testing conditions explain cyber charter schools’ failing grades? American Journal of Distance Education, 33(1), 46-58.
• Berliner, D. (2006). Our impoverished view of educational research. Teachers College Record, 108(6), 949-995.
• Betz, N. (1994). Self-concept theory in career development and counseling. The Career Development Quarterly, 43(1), 32-42.
• Beuermann, D. & Jackson, K. (2019). The short and long-run effects of attending the schools that parents prefer. National Bureau of Economic Research Working Paper No. 29420.
• Black, E., Ferdig, R. & DiPietro, M. (2008). An overview of evaluative instrumentation for virtual high schools. American Journal of Distance Education, 22(1), 24-45.
• Black, D. & Smith, J. (2008). Estimating the returns to college quality with multiple proxies for quality. Journal of Labor Economics, 24(3), 701-728.
• Bryan, J., Farmer-Hinton, R., Rawls, A. & Woods, C. (2018). Social capital and college-going culture in high schools: The effects of college expectations and college talk on students’ postsecondary success. Professional School Counseling. doi:10.5330/1096-2409-21.1.95
• Chingos, M. (2013). Questioning the quality of virtual schools. Education Next, 13(2).
• Conley, D. T. (2007). Redefining college readiness. Eugene, OR: Educational Policy Improvement Center.
• Curtin, R., Presser, S., & Singer, E. (2000). The effects of response rate changes on the index of consumer sentiment. Public Opinion Quarterly 64: 413–428.
• Darling-Hammond, L. (2007). Race, inequality, and educational accountability: the irony of ‘No Child Left Behind’. Race Ethnicity and Education, 10(3), 245-260.
• Ericson, P. & Gardner, J. (1992). Two longitudinal studies of communication apprehension and its effects on college students’ success. Communication Quarterly, 40(2), 127-137.
• Fenning, B. & May, L. (2013). “Where there is a will, there is an A”: examining the roles of self-efficacy and self-concept in college students’ current educational attainment and career planning. Social Psychology of Education, 16, 635-650.
• Ferguson, R. (2012). Can student surveys measure teaching quality? Phi Delta Kappan, 94(3), 24-28.
• Gaertner, M. & McClarty, K. (2015). Performance, perseverance, and the full picture of college readiness. Educational Measurement Issues and Practice, 34(2), 20-33.
• Gecas, V. (1982). The self-concept. Annual Review of Sociology, 8, 1-33.
• Geiser, S. & Santelices, M.V. (2007). Validity of high-school grades in predicting student success beyond the freshman year: High school record vs. standardized tests as indicators of four-year college outcomes. University of California, Berkeley, Center for Studies in Higher Education. Retrieved from https://files.eric.ed.gov/... .
• Greene, J. & Paul, J. (2022). Using parental satisfaction to evaluate virtual charter schools. Educational Freedom Institute Working Paper No. 05-2022. Retrieved from https://efinstitute.org/… .
• Goodwin, B. & Hein, H. (2016). Research says/the x-factor in college success. Educational Leadership, 73(6), 77-78.
• Goyder, J. (1987). The Silent Minority: Non-respondents on sample surveys. Oxford: Polity Press.
• Hannah, M. (2020). Cyber charter enrollments are surging. School districts are picking up the tab. Philadelphia Inquirer. Retrieved from https://www.inquirer.com/… .
• Hawken, L., Duran, R. & Kelly, L. (1991). The relationship of interpersonal communication variables to academic success and persistence in college. Communication Quarterly, 4, 297-308.
• Heckman, J. & Kautz, T. (2012). Hard evidence on soft skills. Labour Economics, 19(4), 451-464.
• Hitt, C., McShane, M. & Wolf, P. (2018). Do impacts on test scores even matter? Lessons from long-run outcomes in school choice research. American Enterprise Institute. Retrieved from https://www.aei.org/… .
• Hooley, T. & Dodd, V. (2015). The economic benefits of career guidance. University of Derby.
• Hout, M. (2012). Social and economic returns to college education in the United States. Annual Review of Sociology, 38, 379-400.
• Hughes, K. & Karp, M. (2004). School-based career development: A synthesis of the literature. Institute on Education and the Economy, Teachers College, Columbia University.
• Indeed Editorial Team (2020). Top 11 skills employers look for in job candidates. Indeed. Retrieved from https://www.indeed.com/… .
• Kamenetz, A. (2020). Virtual charter schools are booming, despite a checkered reputation. National Public Radio. Retrieved from https://www.npr.org/… .
• Kingsbury, I., Beck, D. & Bradley-Dorsey, M. (2022). What kind of students attend cyber schools? Pandemic enrollment as evidence of negative selection. Front Educ. 7:901319. doi: 10.3389/feduc.2022.901319
• Kurlaender, M., Reardon, S., & Jackson, J. (2008). Middle school predictors of high school achievement in three California school districts. Santa Barbara, CA: California Dropout Research Project.
• Litke, D. 1998. Virtual schooling at the middle grades: A case study. Journal of Distance Education, 13(2): 33–50.
• Los Angeles Unified School District (n.d.). College & Career Readiness Survey. Retrieved from https://achieve.lausd.net/… .
• Maranto, R., Beck, D., Clark, T., Tran, B. & Liu, F. (2021). Choosing cyber during COVID. Phi Delta Kappan, 103(1), 30-33.
• McCroskey, J., Booth-Butterfield, S. & Payne, S. (1989). The impact of communication apprehension on college student retention and success. Communication Quarterly, 37(2), 100-107.
• McKillip, M., Godfrey, K. & Rawls, A. (2013). Rules of engagement: Building a college-going culture in an urban school. Urban Education, 48(4), 529-556.
• Molnar, A. (Ed.), Miron, G., Barbour, M.K., Huerta, L., Shafer, S.R., Rice, J.K., Glover, A., Browning, N., Hagle, S., & Boninger, F. (2021). Virtual schools in the U.S. 2021. Boulder, CO: National Education Policy Center. Retrieved from http://nepc.colorado.edu/… .
• Moretti, E. (2004). Estimating the social return to higher education: evidence from longitudinal and repeated cross-sectional data. Journal of Econometrics, 121, 175-212.
• Nasir, R. & Lin, L.S. (2013). The relationship between self-concept and career awareness amongst students. Asian Social Science, 9(1), 193-197.
• Oreopoulous, P. & Salvanes, K. (2009). How large are returns to schooling? Hint: Money isn’t everything. National Bureau of Economic Research. Working Paper No. 15339.
• Paul, J. & Greene, J. (2022). Investigating the relationship between negative selection into online schooling and achievement growth. Education Reform Faculty and Graduate Students Publications. Retrieved from https://scholarworks.uark.edu/edrepub/133
• Paul, J. & Wolf, P. (2020). Moving on up? A virtual school, student mobility, and achievement. (EdWorkingPaper: 20-309). Retrieved from Annenberg Institute at Brown University: https://doi.org/10.26300/1h20-nk64
• Psacharopoulos, G. & Patrinos, H. (2018). Returns to investment in education: A decennial review of the global literature. World Bank Group. Policy research working paper no. 8402.
• Ravitch, D. (2010). The death and life of the great American school system: How testing and choice are undermining education. New York, NY: Basic Books.
• Roderick, M., & Nagaoka, J. (2005). Retention under Chicago’s high-stakes testing program: Helpful, harmful, or harmless? Educational Evaluation and Policy Analysis,27, 309– 340.
• Rubin, R. & Graham, E. (1988). Communication correlates of college success: An exploratory investigation. Communication Education, 1, 14-27.
• Sivin-Kachala, J. & Bialo, E. (2009). IESD Comprehensive Technical Report. Evaluation of the social skills of full-time, online public school students. Retrieved from https://www.k12.com/… .
• Shavelson, R., Hubner, J. & Stanton, G. (1976). Self-concept: validation of construct interpretations. Review of Educational Research, 46(3), 407-441.
• Singer, E., van Hoewyk, J., & Maher, M. P. (2000). Experiments with incentives in telephone surveys. Public Opinion Quarterly 64: 171–188.
• Smith, W. (2008). Does gender influence online survey participation? A record-linkage analysis of university faculty online survey response behavior. Published through Education Resources Information Center.
• Splitzberg, B. (1983). Communication competence as knowledge, skill, and impression. Communication Education, 32(3), 323-329.
• Swann, W., Chang-Schneider, C. & McClarty, K. (2007). Do people’s self-views matter? American Psychologist, 62(2), 84-94.
• Taskinen, P., Schütte, K. & Prenzel, M. (2013). Adolescents’ motivation to select an academic science-related career: The role of school factors, individual interest, and science self-concept. Educational Research and Evaluation, 19(8), 717-733.
• Wats, M. & Wats, R.K. (2009). International Journal of Learning, 15(12), 10-14.
• Wilkerson, D. J., Manatt, R. P., Rogers, M. A., & Maughan, R. (2000). Validation of student, principal and self-ratings in 360″ feedback® for teacher evaluation. Journal of Personnel Evaluation in Education, 14(2), 179–192
• Wladis, C. & Samuels, J. (2016). Do online readiness surveys do what they claim? Validity, reliability, and subsequent student enrollment decisions. Computers & Education, 98, 39-56.