Developing employability skills in vulnerable youth: Designing logic model framework and outcome evaluation using quasi-experiment

1. Introduction

Changing labour markets, economic transformations and new technologies are affecting economic sustainability and disrupting the organisational landscape, while also creating more opportunities and job creation in certain sectors [30]. This demands a higher level of competencies from individuals and they are expected to continuously change to remain competitive [5]. In this context skills development of the youth gains significance. The present skills landscape in India, despite the sufficient manpower isn’t encouraging. India’s Skill Development initiative to skill one crore youth under Pradhan Mantri Kaushal Vikas Yojana (PMKVY) by 2020 achieved approximately 36 lakh people enroled by the end of 2018 [7]. Despite an increase in the number of Industrial Training Institutes (ITIs), only 15.3% were enroled in vocational training with the enrolment rate amongst rural students at 24% and 8.3% amongst urban students (NSS 75th Round).

Skills development gains greater significance in the face of rising unemployment. Over the last few years there has been an improvement in the number of skilled labour from 34% in 2014 to 47% in 2019 but a simultaneous decrease in those who are able to find a job upon completion from 50% to 30% (India Skills Report, 2019). Participatory Labour Force Survey report (PLFS 2018) states that unemployment rates for urban youth (15 to 29 years) was 23.7% and this figure is not far from CMIE (first quarter 2019) unemployment rate of 37.9% in urban youth (20 – 24 years old). Of those surveyed under PLFS 2018, only 2% mentioned having received any formal training and 8% mentioned having some informal training, these figures also align with the Twelfth Five Year Plan that less than 5% of those between 15 and 29 years were formally trained. However, PLFS data reveals that 33% of those formally trained remain unemployed and this highlights the problem of sustainably developing a skilled labour force.

The model of vocational training within the educational institutes in India has so far proved ineffective mainly due to a mismatch between demand and supply of skilled workforce. Vocational Education and Training (VET) and general education have been functioning as two disjoint vertices and thus have been largely inadequate in either providing “work ready” individuals or even upgraded skilled trainers (Jain, 1992; Agarwal et al., 2014; [10], [11]). Within the formal structure of skills development itself, acquired skills of ITI/ ITC graduates do not meet the demands of the industry due to redundant curricula, poor quality in terms of infrastructure and inadequate teacher capabilities [37]. The institutes lack close links with industry and understanding of employers’ needs. A major challenge throughout programs lies in effectively testing the outcomes of such programs in terms of skill development. To add to this, inequalities in opportunities and human development restrict development of potential and aspirations in the vulnerable youth and “manifest as inequalities in outcomes in adulthood” ([15], p. 1).

Vulnerable youth face structural barriers due to poor human capital endowments, lack of relevant skills, lack of mentorship or protective factors [36], thus are likely to be absorbed into the labour market at an early age (ILO, 2015). Most of them cannot afford to remain unemployed due to their unstable economic backgrounds and thus seek work as casual wage labourers in low or middle skilled occupations characterised by low labour productivity (Government of India: 2006; [38]). In this context, the ability to attain sustainable employment implies more significance than gaining employment, where individuals could simply take up lower-level jobs due to financial constraints ([24], pg. 2), and thus can provide an incomplete picture of what the individual has gained [16].

The channel of skills acquisition in vulnerable and at-risk youth is highly informal making them susceptible to poor employment conditions, low productivity and lower wages [32]. There is thus a greater need to create a framework for training and skills development that “suits their economic compulsions” (Government of India, 2016, pg. 13).

Coronavirus (COVID-19) is causing massive disruptions in economic activities, incomes, and work; with rapidly changing labour markets and new technologies, markets demand higher competencies, practical and transferable skills from individuals in order to remain competitive (Bennet, 2006; [27]). Insufficient technical infrastructure to support effective distance learning leads to challenges in acquisition of practical skills and causes disruptions in smooth continuity of TVETs [27]. Disruptions in education, training and skills development are widening pre-existing gaps in access and outcomes; putting the vulnerable youth at risk of falling further behind. With increasing economic hardships, there is a need to expand availability of accessible training programs that can develop better resiliency and maintain engagement in the face of future crises [27].

2. Theoretical framework

2.1. Some existing frameworks of employability and skills development and quasi-experimental designs in education

Studies done on employability and its determinants, have attempted to present frameworks as a guide towards successful movement of students to the labour market. Thus, “learning and employability frameworks” have been provided to help various stakeholders to understand and develop policies according to industry requirements [16], [48]). These focused on providing a learning environment or an ecology consisting of “learners, learning environment and study repository” that influences learning and therefore enhances employability [39]. Model based on five key elements of higher education to achieve “optimum level of employability” as discussed by [4] includes i) disciplinary skills, ii) disciplinary content knowledge, iii) workplace awareness, iv) workplace experience, and v) generic skills. The USEM model is based on four interrelated employability components of i) understanding: subject knowledge, ii) skills: specific and generic, iii) efficacy beliefs: self-awareness and iv) metacognition: self-reflection and regulation [48].

In order to leverage the efforts of different stakeholders, the U.S. Department of Education, guided by CTE, has developed an ‘Employability Skills Framework’ listing a set of general cross cutting abilities that are required to be ‘career ready’ including workplace skills such as technical use and resource management; applied academic skills and critical thinking; and interpersonal skills [41].

The theoretical model of the Career EDGE framework of employability that is widely used focuses on developing subject knowledge and skills, both as a motivator to attain higher education and also to get wider access to employment opportunities [9]. Other than subject knowledge, generic skills such as creativity, adaptability, resiliency, willingness, communication, time management, attention to detail and use of new technologies are amongst those that employers place a higher value on (Harvey et al., 1997). Goleman in his book ‘Working with Emotional Intelligence’ (1998) strongly supports including emotional intelligence in employability models and sets out a framework of emotional intelligence (EI) that reflects how an individual’s potential for mastering the skills of “Self-Awareness, Self-Management, Social Awareness, and Relationship Management translates into on-the-job success” ([20], p. 1); the model gains significance in the current knowledge based economy (Moynagh & Worsley, 2005; [16]).

An important conceptual framework is the Skills Towards Employment and Productivity (STEP) focusing on building right technical, cognitive and digital skills and OJT [44]. Similarly, the Demand-Driven Training Toolkit (DDT) provides research-based and practically applied frameworks that aims to narrow the gaps between what the individual learns through formal education systems and what employer needs (DDT for youth employment- Toolkit, 2018).

2.2. Challenges in building framework for skills development

Disruptions caused by COVID-19 exposes students from low-income households to a greater risk of premature termination of their education and learning opportunities significantly affecting their feeling of self-worth and sense of belonging [40]. Identifying target groups, identifying and matching gaps in skills, co-creating or modifying curriculum, mentorship requires aligning the intervention with employers’ requirements [17]. First, gathering labour market information (LMI) or data on ‘demand for’ and ‘supply of’ labour requires careful selection from national and local surveys, real time market data and/ or through surveys, interviews with identified students, schools, employers and industry experts. Data-based information provides a more accurate and comprehensive foundation for demand and supply analysis while also keeping track of emerging skills.

Second, aligning different stakeholders on the problem, resource allocation, purpose and strategy can be challenging. Often, there is a narrow focus on factors that lead to lack of skills by stakeholders [14]. Developing ‘job-roles’ or skills in a specific sector requires a comprehensive and coordinated approach such as ‘industry-education partnerships’ which are more likely to reach outcomes. Lack of clarity on roles and responsibility, lack of communication and information between various stakeholders leads to imbalances and creates difficulties in reaching the outcomes [18].

Third, the process of selection is challenging. Selecting at-risk, vulnerable youth requires an extensive analysis on multiple personal and environmental elements such as: i) education and skills, ii) access to services, iii) support from family and iv) peer relations [12], [33].

Fourth, imparting practical skills is a key element of vocational training. Recent disruptions have caused challenges in delivery and measurement of practical skills thus it becomes challenging to ensure that students have access to technical infrastructure, devices, connectivity or even uninterrupted electricity to maintain continuity (ILO, 2020).

Fifth, monitoring on-going intervention and outcome evaluation is a challenging process. Changes in industry requirements require adaptable, ongoing and accurate measures. ‘Evidence-based’ results require adequate resources and clarity in methods [19], [45]. Evaluations must be benchmarked against statistically proven direct and indirect outcomes since continuous monitoring and outcome evaluation can be challenging for individuals from vulnerable social backgrounds given the complexities of barriers, they may face in order to skill themselves.

2.3. Why logic model?

There is a greater value in strengthening skills development programs designed for vulnerable youth who live in poverty. This demands different methods of evaluating performance and testing credibility. The complexity of challenges that the vulnerable youth face as well as gaps in the existing education and training frameworks require a tailored approach that sustainably connects such youth to employment opportunities and evaluates such outcomes. To establish program effectiveness, it is observed that the focus is primarily on outcome data and less on what ‘happens during the program to understand the changes in outcomes’ (Martinek, 2017); for example, testing if the participants are motivated, are provided opportunities in decision making. In this context, developing an evaluation process based on logical reasoning ensures continued modifications, looking into ‘what goes on’ during the program and thus avoids a ‘black box approach’ (Patton, 1997). Thus, interventions based on a logic model framework will allow monitoring what intervention ‘is doing’ and ‘is not doing’.

Logic models have traditionally been used as important tools in ‘building community capacities’ and ‘strengthening community voice’ ([47], p. III). The emphasis on providing ‘evidence based’ conceptual framework to ‘maximise the impact of educational investments’ and clearly show a path from investments to impact is discussed in order to build leadership capacity in students (Daughert et al., 2017, p. 4). Generally, logic models have been used as effective action-orientated tools for program planning, identifying outcomes and providing stakeholders with a clear road map ([28]; Martinek, 2017).

The rationale for creating a skills development program for vulnerable youth within the framework of logic model was that first, it will create a conscious process with clear understanding of challenges, inputs, activities and outcomes; second, it will align ‘planned work’ with ‘intended end results’ such as outputs and impact; and third, it will serve as an effective evaluation tool. For this study ‘vulnerable youth’ is defined as individuals who faced constraints in receiving quality education or were possibly first-generation learners, were belonging to low-income households with negligible assets, were either highly likely to drop-out of their existing education (“at-risk”) or had existing gaps in education which may create obstacles in securing employment that would provide adequate wage and social cover such as health insurance. In this context, this study proposes a ‘Employability and Skills Development Logic Model Framework’ as the blueprint of a pilot intervention conducted over a period of 5 months from December 2020 to May 2021 in New Delhi and Mumbai; with the main objective of developing role-based technical skills in vulnerable youth and subsequently smoothen the transition from school to work. Here, two role-based technical skills are identified from the banking, financial services and insurance (BFSI) sector as: 1) microfinance associate and 2) junior data analysis associate (see notes for details).

The main purpose of creating the framework was to introduce a practical model of developing employability skills in high potential youth and investigate the outcomes through quasi experimental research whether or not the logic model increases employability skills amongst such youth. For this study, ‘employability’ is defined as a set of skills and attributes that make an individual more likely to secure and maintain an occupation, be satisfied and successful in it [9], [24]. Nine measurement scales for three domains testing employability were developed: work-based domain, interpersonal domain and behavioural domain.

Framework emphasis that training interventions should not only be means of getting a job, but should be able to take the edge off social disadvantages of exclusion, discrimination or even addiction and personal disadvantages such as low sense of self-efficacy and lack of future identity ([3]; Mangoche, 2014). Framework is adapted from ‘Kellogs Foundation Guidelines for Developing Logic Model’ and is directly aligned with Target 4 under the Sustainable Development Goals (SDGs), and specifically aligned towards Target 4.3 which is to ‘Ensure equal access to affordable and quality technical, vocational and higher education’; Target 4.4.1 which is to ‘Substantially increase the number of youth and adults with ICT skills’.

3. Research questions and framing the hypothesis

It was hypothesised that there would be significant differences in employability of vulnerable youth with or without implementation of logic model framework designed to develop role-based demand driven skills. The focus first is to test in practice the validity of framework as a logic model for designing and testing an intervention for vulnerable youth. Furthermore, in order to develop employability, three skill domains are identified and development of these skills are embedded within the key activities of the logic model framework. The three broad parameters for developing employability were established as: i) work-based, ii) interpersonal and iii) behavioural. To test the hypothesis, the following research questions were proposed and analysed.1)

Does integrating work-based learning in curriculum enhance graduate employability amongst the students of the experimental group more than those in the comparison group?2)

Do experimental group students perceive an effective advancement in their interpersonal skills as compared to those from comparison groups?3)

Did the intervention impact positive behavioural attributes amongst experimental group students as compared to comparison group students?

4. Methodology

4.1. Subjects

For this study, of the 387 students mobilised, two groups were formed: a comparison group and an experimental group. Mean age of both groups is 21 years. More than half (57.4%) have annual household incomes of less than Rs. 70,000. Around 40% are first generation learners. In order to control for selection bias and student variability, the same set of selection criteria, interview questions and interviewers were used to select participants for the two groups. Participants were tested on the basis of four broad parameters i) academic performance and orientation (secondary/ higher secondary/ percentages or scores), ii) household income and iii) demographic backgrounds (age, education of parents, earning and dependant members) and iv) commitment towards participating in the program.

For the last parameter, a fully refunded deposit fee of Rs. 5000 was taken which was returned on completion of the program. A total of 76 students were selected as the experimental group of which 65 joined the program. Of the remaining participants, 65 were reached out to who showed willingness to take all survey questions administered post intervention. This group was the comparison group. These students were not subject to any vocational & technical intervention or training. Comparison group at baseline consists of 65 students and post intervention only 57 participants remained, eight students did not respond to questionnaires nor were available for calls. It is important to emphasise here that quasi-experimental research can be successfully run only on the basis of carefully matched groups, in the absence of which the evaluations can be incorrect at worst and misleading at best [11], [22].

4.2. Development of domains and measurement scales

A baseline survey consisting of three pre-tests were administered to the comparison group (65 students) and experimental group (65 students). These pre-tests were designed around factors that influence employability and development of skills. Of-course learning sector specific or work-based skills has to be the central concept in the framework. There remains less argument that the main motivator to attain a degree or higher qualification is to secure a better employment opportunity; better qualified have far greater job opportunities (Johnes, 2006). However, in vulnerable groups, over and above the need to develop sector specific skills, the significance of improving “generic skills” also called ‘‘transferable skills’ is in fact higher. Generic skills are “prime qualities” that can enhance learning across disciplines ([4]; Knight & Yorke, 2002) and are highly desirable by employers (Harvey et al., 1997). Some of the key generic skills include: responsibility, willingness to learn, team work, management and independent work, time management, communication, coordination and organisation, written and numeracy skills (The Pedagogy for Employability Group, 2004). Lastly, the study perceived that creating ‘self-efficacy’ or a sense of self-worth will be the real crucial trait that allows vulnerable students to execute their own course of action when required and keep them motivated. Based on these factors influencing employability, three domains were created to test groups on work-based, interpersonal, and behavioural skills. Three broad parameters of testing skills within that domain is as follows:A

Work-based: i) Applied knowledge ii) Technological skills and iii) Information use.B

Interpersonal: i) Effective communication skills ii) Personal presentation, and iii) Time management.C

Behavioural: i) Motivation and confidence ii) Future identity and iii) Willingness, responsibility and self-discipline.

In order to measure employability skills in students of experimental and comparison groups before the intervention, a 5-point Likert scale questionnaire was drafted and administered. The questionnaire consisted of a total of 20 questions: nine in domain A, six in domain B, and five in domain C. Scores were assigned to each question to create a baseline index in the following manner: 5 = “strongly agree/ always/ excellent”; 4 = “somewhat agree/ often/ good”; 3 = “neither agree nor disagree/ neutral/ average”; 2 = disagree/ rarely/ poor”; and 1 = “strongly disagree/ never/ very poor”, such that a minimum score of 20 and maximum score of 100 is created. Skewness-kurtosis test of normality to check the reliability of the questionnaire and to ensure data followed a normal distribution was used. In this study, the procedure for quantitatively developing measurement scale for each domain is as follows.

4.2.1. ‘Work-based’ scale

In order to measure students’ work-based skills, a total of nine sub-parameters were drafted based on the studies of Bennett et al., [4], Martin (2004) and Martin (2014). Out of the list of five elements ‘ensuring optimum level of employability’ by Bennett et al., [4] and 14 employability skills as identified by Martin et al., [35]; this study condensed to the following nine: 1) technical know-how, 2) job-role/ content know-how, 3) workplace/ job-specific, 4) sector specific awareness, 5) academic assistance, 6) placement readiness, 7) preference on quality over ‘any job’, 8) literacy and numeracy, 9) information use. Full score totalled 45, with a minimum score of 9. A questionnaire on work-based domain was administered to both groups pre-intervention. Average score of the comparison group was 19.71 with standard deviation (SD) of 3.60. Average score of the experimental group was 19.42, with a SD of 3.42. Thus, both groups performed similarly under this domain. This is also referred to as a baseline survey.

4.2.2. ‘Interpersonal’ scale

In order to measure students’ interpersonal skills, a total of six sub-parameters were drafted based on studies of Bennett et al., [4] and Nair et al., (2009). This study identified: 1) communication, 2) non-cognitive/ analytical, 3) resource management, 4) time management, 5) personal presentation and 6) teamwork/ or ability to work independently. The scale scores ranged between 30 and 6. To establish the validity of questions of interpersonal scale, a series of interviews with students of experimental and comparison groups were conducted. In addition, an independent-samples t-test is used to compare the scores of both groups pre-intervention. A questionnaire with interpersonal domain was administered to both groups pre-intervention. Average score of comparison group was 13.62 with a SD of 3.15. Experimental group’s average score was 13.5, with a SD of 2.76.

4.2.3. ‘Behavioural’ scale

In order to measure behavioural skills of the students of both groups a total of five sub-parameters that are incorporated from Career EDGE model, Bandura’s (1995) “Ss” and Fugate et al. [19], [45] are: 1) self-efficacy/ motivation, 2) future identity, 3) willingness to learn, 4) responsibility and 5) self-discipline. Behavioural scale incorporates the impact of vulnerable students’ social background and support system that influences their emotional well-being. The scores ranged from 25 to 5. Average score of comparison group was 13.06 with a SD of 2.01. And that of the experimental group were 13.55 and 2.07 respectively, thus, it was established that the groups were similar before intervention. (See table 1). The two groups are tested on these three domains after the completion of training for the experimental group under the quasi-experimental analysis. Findings for post-intervention are given later under the results section.

Table 1. Measurement Scale Summary: Experimental and Comparison groups pre intervention.

Empty Cell	Work Based	Interpersonal	Behavioural
Empty Cell	Mean	SD	Mean	SD	Mean	SD
Experimental Gr Baseline	19.42	3.42	13.54	2.76	13.55	2.07
Comparison Gr Baseline	19.71	3.60	13.62	3.15	13.06	2.01

4.3. Quasi-Experimental research

A quasi-experimental research exercise was conducted for a period of five months. The research procedure was as follows (see Fig. 1).

5. Framework

The Employability and Skills Development Logic Model Framework has 5 main blocks: i) Inputs, ii) Outputs iii) Intermediate Outcomes, iv) Outcome Evaluation and v) Impact. The process is cumulative, elements in each block are based on previous course of action or stage. Data is collected by quantitative and qualitative methods at three levels: (i) Level 1 (Baseline Survey) during planning and identifying collaborative partnerships phase. (ii) Level 2 (Assessment) during activities/ corresponding activities phase and (iii) Level 3 (Evaluation) for outcome evaluation. Methodology for three levels is designed and required indicators are identified during the initial stages of the pilot (also see methodology section). Description of each block contains a rationale (why was the block included), data and methodology (how information was collected and analysed) and results (what was achieved). Here’s what the framework looks like (see Fig. 2).

5.1. Inputs

‘Inputs’ include “resources and infrastructure” that will be required to run the program [8]. Assessment and Planning (What) includes setting objectives and purpose of work. The main objective of the program is to train vulnerable and at-risk youth in trade specific skills and subsequently transition them into sustainable employment. The purpose is to establish that skills development in vulnerable youth requires a demand driven approach that is tailored to their needs (DDT Toolkit, 2018). Constant changes and disruptions brought by technology, globalisation, education and even pandemic recently, have led to changes in how work is done and how it is organised. It is thus essential to keep track of ‘skill shortages’, what jobs will remain and what will emerge and ‘disappearing-jobs’. Labour Market Information or LMI is an essential first step in evaluating sectoral changes and requirements to identify skills the youth need to develop and align them with their needs and capacity to learn as well as the trainers forte to train. LMI was essentially administered for the BFSI sector in India, where the rate of change is exponential causing light skill redundancies and thus the need for investments in reskilling and upskilling. BFSI is a high growth sector, with projected demand of 8.5 million labour force by 2022 (National Skill Development Corporation), with Maharashtra and Uttar Pradesh generating the highest number of skilled and employable youth [25], [26]. LMI also showed that the BFSI sector is facing an employable manpower challenge with very few vocational courses aligned to job roles therefore investments aligned to reskilling and upskilling will be required. Jobs in this sector will be divided mainly into: data management, risk management and alliances.

5.1.1. Level 1 (Baseline survey)

is used for i) selection of experimental and comparison groups, ii) identifying skill gaps, iii) identifying job relevant skills and iv) designing course curriculum according to current or future requirements of the industry. Selection criteria for the total sample group is a set of predetermined demographic factors such as household income, education and employment of parents, education type, number of gap years to identify the extent of vulnerability faced.

Identifying skill gaps required us to collect data through online questionnaires circulated amongst 65 experimental group students and 19 selected employers from the BFSI sector. All respondents were given a list of seven skills based on comprehensive study by Martin et al. [35]. The study identified 14 employability skills and analysed these in order of priority and importance, devoid of these skills could prevent a candidate from employment [42]. The seven selected of these were: Literacy and numeracy skills, general Information and technology, communication, time keeping, team-work, advanced vocational and role-based skills, and personal presentation. Respondents were asked to rank these on a scale of 1 to 7. Rank 1 to the skill they perceive as most important to attain employment, and rank 7 to that considered least important. Results in Tables 2 and 3 show a mismatch between perceptions of the students and the employers.

Table 2. Mean scores of skills perceived as important by students ranked from 1 to 7.

Skills	Literacy & numeracy	General IT	Comm.	Time-keeping	Team-work	Advanced Vocc. & role-based	Personal presentation
Means	2.48	1.81	5.58	4.04	5.23	2.01	6.86
SD	0.87	0.7	0.63	0.71	0.95	1.32	0.39

Table 3. Mean scores of skills perceived as important by employers ranked from 1 to 7.

Skills	Literacy & numeracy	General IT	Comm.	Time-keeping	Team-work	Advanced Vocc. & role-based	Personal presentation
Means	2.22	2.11	2.11	5.66	6.88	3.55	4.55
SD	0.83	1.17	1.76	0.71	0.33	0.73	0.73

The skills perceived by students as least important (mean scores >5.5) were ranked moderate to high importance (mean scores < 2.5) by employers. Findings suggest that while the students could be aware of employability skills in general, they were not able to rank the skills as per demand in the labour market. In particular, skills such as communication and personal presentation identified as least important by students were prioritised by most of the employers. Surprisingly, general IT skills and literacy and numeracy were identified by both groups as highly important, notwithstanding that these students regardless of their vulnerable backgrounds, had performed fairly well academically at school level suggestive of high priority given to literacy and numeracy.

5.1.2. Participants of different stakeholders

(Who) strengthen outcomes such as well-designed customised curricula, mentoring, work-readiness and hiring. Partnerships benefit all stakeholders building exposure and awareness of the work environment. The pilot provided in-class learning conducted by educational institutes, teachers from universities were selected for training the students. Mentors were assigned to constantly guide, motivate or provide any emotional support. On-the-job training was provided by private companies operating in the BFSI sector.

5.1.3. Activities

(How) involve tailoring and modifying existing curriculum according to the needs of the employers and thus, co-designing with employers and their expectations of the job roles. Two trade specific job roles are identified: 1. Microfinance Associate, 2. Data Analysis Associate. These two roles have a steady job growth in future that vulnerable youth otherwise find difficult to secure. It was a given assumption within the stakeholder groups that students from the vulnerable groups do not learn technical skills of higher order well and thus cannot be made employable for data analysis. Selecting this job role was to establish that students from vulnerable groups could aspire for higher order skills and have the aspiration to upskill themselves for future work opportunities. The microfinance job role was identified because jobs could be made available at entry level and students could seamlessly transit to jobs while still studying.

Training methods included a combination of in-class and workplace-based learnings. Mentoring is an instrumental element of skill development in vulnerable groups, improving both cognitive and non-cognitive skills ([13], pg. 32). Two mentors were assigned for each group in New Delhi and Mumbai that were sensitised on the vulnerabilities and background of the students. Capacity building included selection of 48 teachers and providing professional training for one week (total of ten hours). Post training teacher’s assessments were conducted on four parameters: i) communication, ii) digital knowledge, iii) subject knowledge and iv) responsiveness to needs of students. See Appendix for summary and data description.

5.1.4. Level 2 (Assessment)

A mixed-method assessment for the experimental group is conducted that includes: standardised summative tests, class presentations and computer-based assessments (CBA). All assessments were credit based and a composite credit was generated for each student based on their performance throughout the training program. Alongside feedback is taken from teachers and mentors to identify changes in performance of students and the challenges they face during module delivery; adding emphasis that interventions must be continuously monitored regardless of evaluation ([13], pg. 20). Assessment did not require a pre-test and post-test design since the students were previously not trained for sector specific skills.

5.2. Output

The ‘logic’ in the logic model is realised when the inputs and intermediate outcomes result in desired outputs. Outputs brings out the ‘change’ section of the framework. Here, elements that can provide observable evidence of change during the course of the intervention are classified as outputs, and can be considered as pre-conditions that are required to be met in order to achieve the outcomes. As part of the intervention, outputs were categorised as those pre-conditions that can be directly linked to the activities, these include providing a smooth transition from school to workplace with sustainably employing the youth and providing professional training to 48 teachers selected as one of the participants during the ‘activities’ phase. Furthermore, indirect evidence of activities are those pre-conditions that are hypothesised to arise due to the intervention however may not necessarily be ‘caused’ due to the intervention for the experiment group. Gap between skills necessary for employment and ‘actually’ securing employment is narrowed and there is improvement in workforce readiness, these are observable milestones of the intervention but cannot be directly stated as occurred because of the activities. Of course, in interventions that do not accommodate for sustainable long-term placements, the experiment group will benefit from training and mentorship thus leading to an improvement in workforce readiness and enhanced skill set, however may not result into gainful employment immediately.

5.3. Intermediate outcomes

Identification of outcomes is the first step while setting up of the framework. Vulnerable youth often not only lack crucial trade specific skills, there seems to be a lack of self-efficacy and sense of future identity, therefore allowing such factors that could change as a result of intervention aimed at improving employability amongst such youth. Intermediate outcomes included development of trade specific and interpersonal skills, developing sense of future identity and creating inclusiveness by changing exposure and consequences for the vulnerable youth. Working backwards from such outcomes (also known as reverse logic), it is essential to identify what are the pre-conditions required to achieve these outcomes. The outcomes provided in this framework directly address key factors that limit opportunities of sustainable employment amongst the vulnerable and at-risk students. The outcomes are achieved only once a set of differently identified outputs are met [31].

During the first phase of selection, during the interview a majority of students showed a high interest in learning skills and attain sustainable employment with a sole objective of contributing to family income. This information was shared by the maximum number of students being interviewed. In vulnerable youth such motivations result in urgent requirements to obtain low-skilled unsustainable jobs. In this context, responding to the urgent need to provide a ‘short term entry’ into the labour market is an incomplete solution. Given the economic and demographic vulnerability of the target group, the pilot emphasised on a learner centric approach ensuring not only a smooth transition from school to work, but also development of analytical thinking, team work, communication skills and sustainable jobs. The pilot provided sustainable employment with a formal contract for 61 of 65 experimental group students in the BFSI sector. The pilot also built technical and digital skills in BFSI aligned courses amongst 48 teachers.

Vocational and technical training are not and should not be a “quick fix” to address the problem of high unemployment amongst such students, therefore limiting themselves to reducing barriers to enter the labour market [6]. Development of role-based technical training along with attention on curating interpersonal skills and mentorship is recognising that training should not only be a means of better employment opportunities or financial improvements but also take the edge off social disadvantages such as social exclusion, lower sense of self-efficacy, addiction, discrimination in students or even early marriage and pregnancy in young women. An effective intervention therefore includes attempts to change exposure and consequences of vulnerable youth as part of its intermediate outcomes.

5.4. Outcome evaluation

The objective of outcome evaluation is to assess the progress and changes in the experimental group caused by the intervention and statistically measure such progress and changes. In order to effectively test the viability of the results, quasi-experimental research is designed and conducted. Details of quasi analysis are given under the methodology section

5.4.1. Level 3 (Evaluation)

includes testing on specific predetermined parameters. Work-based skills and interpersonal skills were the two domains chosen for the comparative study. A third domain on behavioural skills is also added since behavioural changes indirectly correlate with changes in the first two. The broad parameters created in each of the three domains, questionnaire and measurement scales are discussed under the methodology section.

5.5. Impact

Impact of skills development interventions for vulnerable youth needs to be examined in the context of its effect on economic and social outcomes over a longer period of time [29]. The impact is expected in terms of enhanced employability and contribution towards SDG Targets 4.3 and 4.4.1. However, it is crucial to measure whether intervention led to a sustainable change in the consequences for the target vulnerable youth much after training is complete. This pilot intervention therefore proposes to reach out to students of the experimental group after 18 months of completion of the program and conduct an impact evaluation based on three parameters: i) the probability of being currently employed, ii) changes in household financial condition and iii) changes in perception towards work and individual potential.

6. Results

Comparing the two groups post training of the experimental group on the basis of three domains showed differences between the scores across domains. Mean scores were higher for the experimental group against that of the comparison group (table 4).

Table 4. Measurement Scale Summary: Experimental and Comparison groups post intervention.

Empty Cell	Work Based	Interpersonal	Behavioural
Empty Cell	Mean	SD	Mean	SD	Mean	SD
Experimental Gr Post	33.45	2.91	18.58	2.90	20.77	1.49
Comparison Gr Post	21.18	3.70	12.98	3.14	14.28	2.68

On comparing pre- and post-intervention means of the experimental group, the percentage change within group scores was greatest for work-based skills i.e., domain A (72,2%), followed by improvement in behavioural skills i.e., domain C (53.2%) and improvements in interpersonal skills i.e., domain B (37.2%).

Independent-samples t-test is conducted to compare experimental and comparison groups before the intervention (table 5). No significant difference is found between the scores of the experimental group (M = 46.5, SD = 5.16) and comparison group (M = 46.4, SD = 6.03) before intervention; t (128) = −0.109, p = 0.913. Hence the null hypothesis of equality of two groups cannot be rejected.

Table 5. t-test Independent Samples Assuming Equal Variances: Experimental and Comparison groups pre intervention.

Empty Cell	Obs	Mean	SD	[95% Conf. Interval]	df	t Stat	P(|T| > |t|)
Experimental Gr Baseline	65	46.51	5.16	45.22828 47.78711	128	−0.1093	0.9131
Comparison Gr Baseline	65	46.40	6.03	44.90504 47.89496

Note: t-Test Significance Level 5%.

As shown in table 6, significant difference is found between the scores of the experimental group (M = 72.8, SD = 4.87) after training and comparison group (M = 48.4, SD = 6.60); t (120) = −23.36, p < 0.001.

Table 6. t-test Two Sample Assuming Equal Variances: Experimental and Comparison groups post intervention.

Empty Cell	Obs	Mean	SD	[95% Conf. Interval]	df	t Stat	P(|T| > |t|)
Experimental Gr Post	65	72.80	4.87	71.59307 74.00693	120	−23.3688	0.0000
Comparison Gr Post	57	48.44	6.60	46.68642 50.19077

Note: t-Test Significance Level 5%.

In order to test the effectiveness of the training on the treatment group, a paired sample t-test is used to compare the experimental group before and after intervention. Table 7 shows significant differences (showing improvements in scores) between the scores of the experimental group before intervention (M = 46.50, SD = 5.16) and after intervention (M = 72.8, SD = 4.87); t (64) = −31.22, p < 0.001. This implies that the null hypothesis of equality of two groups is rejected.

Table 7. t-test Paired Sample: Experimental group pre (baseline) and post intervention.

Empty Cell	Obs	Mean	SD	[95% Conf. Interval]	df	t Stat	P(|T| > |t|)
Experimental Gr Baseline	65	46.51	5.16	45.22828 47.78711	64	−31.2252	0.0000
Experimental Gr Post	65	72.80	4.87	71.59307 74.00693

Note: t-Test Significance Level 5%.

Due to a small sample, a Skewness-Kurtosis test for Normality is utilised to ensure that the data followed a normal distribution and to strengthen the validity of the t-test [21]. This test accepts the hypothesis of normality when p > 0.05. A non-significant result of p > 0.05 for all sample data sets is observed as mentioned in table 8.

Table 8. Skewness-Kurtosis test for Normality in the groups to determine validity of t-test.

Empty Cell	Skewness	Kurtosis	Adj. chi2	Prob>chi2	Decision
Comparison Gr Baseline	0.0541	0.9640	3.88	0.1436	Retain the null hypothesis
Comparison Gr Post	0.1646	0.5725	2.36	0.3076	Retain the null hypothesis
Experimental Gr Baseline	0.6123	0.8748	0.28	0.8686	Retain the null hypothesis
Experimental Gr Post	0.4225	0.4133	1.36	0.5068	Retain the null hypothesis

Note: Skewness-Kurtosis Test Significance Level 5%. The test accepts the hypothesis of normality when p-value >0.05.

To further evaluate the data, the Wilcoxon Signed-rank test which is suitable to test the null hypothesis of similarity of score rankings of pre- and post-intervention in smaller samples is conducted. First, checking for similarity in experimental and comparison groups pre-intervention, results in table 9 show a non-significant difference, z (65) = −0.376, p > 0.05 (p = 0.70). This implies the null hypothesis of similarity of two samples cannot be rejected.

Table 9. Wilcoxon signed-rank test: Experimental group pre- intervention and Comparison group pre- intervention (n = 65).

Empty Cell	Obs	Sum of ranks	Adj. Variance	z Stat	Prob > |z|
Positive ranks	28	1013.5	23,361.75	−0.376	0.7068
Negative ranks	35	1128.5
Ties	2	3
Total	65	2145

Note: Wilcoxon Signed-rank Test Significance Level 5%.

Table 10 shows a significant difference between the experimental group before and after the intervention, z (65) = −7.011, p < 0.05 (p = 0.00). Comparing the difference between scores of the comparison and experimental groups (after the training was completed for the experimental group), results suggest significant difference between the two groups, z (57) = −6.569, p < 0.05 (p = 0.00). These findings mirror the results of t-tests conducted.

Table 10. Wilcoxon signed-rank test: Experimental group pre-intervention and Experimental group post- intervention (n = 65).

Empty Cell	Obs	Sum of ranks	Adj. Variance	z Stat	Prob > |z|
Positive ranks	0	0	23,400.13	−7.011	0.0000
Negative ranks	65	2145
Ties	0	0
Total	65	2145

Note: Wilcoxon Signed‐rank Test Significance Level 5%.

Results show that while the experimental group has largely benefited from the intervention, a greater indirect outcome has been in the form of increase in self-confidence and motivation. The study therefore lays emphasis on assessing self-efficacy, motivation and sense of future identity as an important component of skills development and employability.

7. Discussion and limitations

The framework presented in this paper served as a useful tool to outline the pilot intervention for developing role-based skills and achieving certain positive behavioural changes in vulnerable youth. It brought clarity in alignment of different stakeholders, data collection and analysis, monitoring and evaluation. A key component of the pilot was to ensure behavioural changes such as increased motivation, self-efficacy and a sense of future identity that was lacking in these students during the initial phases. amongst factors affecting self-efficacy, four can be seen as having implications amongst the students from poor and vulnerable backgrounds, these are: a) past performance and the manner in which past success or failure are interpreted; b) modelling or observation of others with perceived similarities performing same tasks having an influence on self-belief. Modelling is especially useful for vulnerable youth given the observed lack of certainty or confidence amongst them; c) verbal persuasion and providing direct encouragement or discouragement having impact on self-efficacy. Interestingly, discouragement has a greater bearing on reducing a sense of self-worth than encouragement has on increasing it. amongst the vulnerable youth, the role of positive and continuous verbal persuasion as part of behavioural changes gains significance; and d) somatic and emotional state influences self-efficacy where student’s negative emotions can affect their judgement.

Of course, by reducing stress reactions caused by their vulnerabilities, poor youth’s state of self-efficacy can be modified [2], [43]. Even within the Social Cognitive Theory (SCT that started as the Social Learning Theory or SLT by Albert Bandura), self-belief has crucial role of influencing behaviour and motivation and peoples believes in themselves are key variables of personal cultural and social achievement [43]. Self-efficacy is seen as the key factor in influencing academic performance at school and college levels, and poor academic score may not necessarily be due to lack of aptitude but due to lack of belief in themselves. A vulnerable student’s sense of self-worth or personal value is the fountainhead affecting their overall sense of self, how valuable they perceive themselves. And low self-esteem immobilises them from taking newer challenges, responsibility and even handling criticism.

First it was essential to identify the skill gaps and align those with the capacity of the training organisation. During the assessment and planning phase, labour market information (LMI) was specifically conducted for the BFSI sector, which itself is a high growth sector with states like Maharashtra and Uttar Pradesh projected to generate the highest number of employable youths. It was observed within the stakeholder groups that students from vulnerable backgrounds do not learn technical skills of higher order well and therefore job roles such as data analysis associate remain out of reach for them.

Surprisingly, out of all the students mobilised, a majority were ones with a good level of previous academic performance, especially with good numeracy skills despite their vulnerable social backgrounds. The pilot therefore wanted to test (and hopefully prove) that such students could aspire for higher order skills in a high growth tech driven sector such as the BFSI. Microfinance associate was the second job role chosen since few students were still studying at undergraduate levels and therefore an entry level transition into the banking sector could be facilitated without disruption to on-going education.

Second, on identifying the experiment and comparison groups three broad measurement scales were developed to measure overall employability of these groups. The parameters on which scales were built were: work-based, interpersonal and behavioural, which was further grouped into 3 sub-parameters each. Findings of scores of experimental and comparison groups pre intervention showed similarity between the two, thus establishing a reliable foundation to conduct quasi experiment research [1]. It was observed that the mean scores were low across all domains, interestingly both work-based and behavioural remained less than interpersonal; suggesting strong emphasis on developing work and behavioural skills.

Third, identifying skill gaps in terms of skills perceived important by students and by the potential employers, threw light on a few important observations that may be otherwise ignored due to existing assumptions regarding employability of vulnerable youth. While students were aware of employability skills in general, for them, employability would have greater influence on their livelihoods and financial security suggesting that they had clarity while emphasising on skills such as numeracy, role based vocational and general IT. For employers who look for competent employees, the value was placed on well-developed generic skills such as time-keeping, personal presentation and communication. This makes it important for employers to consider that such youth, once trained, continue to have access to reskilling and proactive responses in upskilling attributes that are important to them.

Fourth, using quasi analysis for outcome evaluation indicated significant differences between experimental group pre- and post-intervention strengthening support towards demand aligned framework. Testing for differences between experimental and comparison groups post intervention suggested that the former had overall benefited from the intervention, the latter did not show any significant changes as compared to that observed pre intervention. It can be suggested that while developing employability can be viewed as collective responsibility of multiple stakeholders, the role of educational institutions remains paramount.

Even though the outcomes of the pilot provide strong evidence in support of demand-aligned training, care must be taken to apply the framework into interventions. A number of initiatives take up a simplistic view of training and thus provide a ‘quick-fix’ with inadequate focus on interpersonal skills, behavioural changes or securing sustainable employment. These may seem justified as the need for providing vulnerable youth the opportunities to enter the labour market is greater and more urgent in the context of evolving labour markets. At the same time significance of formal education and job security cannot be overruled by short term informal training or unsustainable jobs. Vocational interventions bring added benefits only on foundations of strong general education, improving productivity and stability of jobs.

This intervention faced reluctance from parents of girls’ students during enrolment, which limited the scope of this work. As a socio-cultural problem this required persuasion, counselling and assurance from teachers and trainers to the parents. The participation of girls remained low therefore further research is needed to understand the implications of such a framework on gender or other exclusive groups. Recruiting experienced teachers and trainers was a major challenge. Some experienced teachers lacked digital teaching skills and had to be trained accordingly. Teachers and trainers had to be sensitised and provided guidelines on delivery methods tailored to the requirements of vulnerable and at-risk youth. Their commitment to train such youth was monitored and constructive feedback was required to keep the trainers motivated and involved in the program. Limitations remain in order to test the long-term impact on the experimental group and needs to be conducted after 18 months post intervention. Research is also required to evaluate scalability of the framework and ability to train and sustainably place vulnerable groups in other sectors in the context of COVID-19 market disruptions. The results of this study are not representative of all sections of vulnerable youth that are at risk of dropping out, therefore further research is needed at the larger scale.

8. Conclusion

This study summarises strategic points for improving employability during skills development interventions and its implications for policymakers and managers.

8.1. Maintain engagement and make it accessible

Vulnerable youth face difficulties and real struggles on a daily basis. Struggles alter their aspirations. Most students mobilised did not have an adequate environment to study at home, many faced challenges in practising technical modules due to lack of laptops or computer systems. Interventions primarily require tailoring around these challenges and struggles. Longer duration programs could struggle with retention. Thus, flexibility in method of delivery, part-time or weekends, involving mentoring, providing financial support and social security, stipend, travel expenses have positive implications [46].

8.2. Ensure quality and make it relevant

Strong multi-stakeholder partnerships are key in delivering demand driven interventions with sustainable outcomes. Partnerships can ensure quality by giving required exposure to the workplace and opportunities for on-the-job training. For employers it gives access to work-ready candidates and improved retention. Framework must incorporate ‘education-industry partnership’ and leverage knowledge of markets and pedagogy to design relevant curriculum aligned with the requirements of the employer.

8.3. Emphasise monitoring and evaluation and make it accountable

Intervention must have a strong monitoring system to be able to continuously track progress regardless of whether the program will be evaluated or not. Effective monitoring generates data on the profile, progress, changes in overall well-being and identifies unintended consequences, positive or negative. Evaluations are less implemented and thus the necessity of rigorous evaluations have been emphasised by academic researchers and policy makers [7].

8.4. Make social outcomes count

Vulnerabilities are difficult to measure. But once it is understood that challenges that vulnerable youth face are long-established, then focusing solely on training is an incomplete solution [23], [34]. Social inclusion is becoming an important element for interventions seeking to improve the overall well-being, break down barriers and develop self-efficacy. Of-course employment itself is visible evidence of social inclusion, but to enable vulnerable youth to be able to make a choice about their role in the labour market, we still have a long way to go.

Notes

1. Counterfactual: Counterfactuals can be determined from the comparison group and not the experimental group. It means what would have happened to the participants in a program had they not received the intervention.

2. Comparison group: A non-randomly designed quasi-experiment required to select students (from the same population) with matched backgrounds as the experimental group. This group did not receive any intervention and is subject to evaluation. They are used as a standard for comparison against the experimental group.

5. Baseline: Data describing the characteristics of students across both experimental and comparison groups before the intervention.

6. Selection bias: Statistical bias between experimental and comparison groups were minimised and thus both samples were chosen from the same population on carefully matched demographic, educational and household-income parameters.

7. Sustainable employment: providing jobs that are secure in terms of social benefit and insurance and medical cover given to the employees. Students are thus placed as formal employees and not contract workers.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors would like to thank Dr Abhinava Tripathi and Dr. Lalitagauri Kulkarni for review and feedback. Thanks to Rahul Ranadive and Vinod Nair at The Skills Development Program, BSE Institute. Mayank Paraswani for software support. Authors take responsibility for all mistakes and errors in this paper. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Appendix 1. Summary of Key Features.

Experimental Group	65 participants
Comparison Group	65 participants (pre-intervention) and 57 participants (post-intervention)
Implementing Agency	BSE Institute, Mumbai
Mobilisation	Multiple channels such as self-help groups (SHGs), direct community outreach at night-schools, National Institute of Open School (NIOS), distance education institutes and local colleges
Group Characteristics	12th Pass; Graduates from low-income; under-privileged backgrounds; likely dropouts due to social and economic vulnerabilities; Age range is 18–25 years
Courses	1) Microfinance Associate and 2) Junior Data Analysis Associate
Course Duration	1) 130+10 hrs and 2) 100+10 hr
Mode of delivery	In-class learning, on-the-job training, on-line modules and class presentations
Mode of assessment	In-class
Number of teachers	48
Hours of teacher training	20 + 7 hrs
Number of mentors assigned	4
Industries partnership	19

Note: Mobilisation mentions the channels through which students were identified. Number of hours are given as hours spent on job-specific technical training and additional hours are for interpersonal training.

Appendix 2: Data Description (N = 387)

Female: 59%; Male: 41%.	No. of dependant members in a household is more than three (> 3) in 60.7%
Rural: 27.9%; Urban: 72.1%	Type of accommodation: Rented home: 55.7%, Own home: 44.3%
Social class division: General: 72.1%; Scheduled Castes: 14.8%; Other Backward Castes: 11.5%; Schedules Tribes: 1.6%	Household vehicle ownership: 37.7% 62.3% had no vehicle
Percent of Students with gap years: 9.3%	Students who perceive vocational training as crucial: 78%, Not sure: 22%.
Annual Household income: Rs. 70,000 or less is 57.4% Rs. 70,000 to 2,73,098 is 29.5% Rs. 2,73,098 and above is 0	Attained some form of vocational training previously 34.4% out of which, students with no employment after training: 90.9%. Short run, one time employment: 9.1% students
No. of earning members in the family: 60.7% had one member; 23% had two members; 9.8% had three members	Students who want to utilise future salary as financial support for family and own education: 100%

Appendix 3: Data Description of Experimental Group (n = 65)

Female: 14; Male: 51	No. of dependant members in a household is more than three (> 3): 0
Rural: 0; Urban: 65	Type of accommodation: Rented home: 30 Own home: 35
Social class division: General: 47; Scheduled Castes: 7; Other Backward Castes: 11; Schedules Tribes: 0	Household vehicle ownership: 33 had no vehicle
Students with gap years: 2	Students who perceive vocational training as crucial: 65
Annual Household income: Rs. 70,000 or less is 41 Rs. 70,000 to 2,73,098 is 24 Rs. 2,73,098 and above is 0	Attained some form of vocational training previously out of which, students with no employment after training: 6 Short run, one time employment: 5 students
No. of earning members in the family: 44 had one member; 21 had two members; none (0) had three or more than three.	Students who want to utilise future salary as financial support for family and own education: 65

References

[1]S.-.G. Back, E.-.H. HwangA Quasi-Experimental Research on the Educational Value of Performance AssessmentAsia Pacific Edu. Rev., 6 (2) (2005), pp. 179-190, 10.1007/bf03025828View article View in ScopusGoogle Scholar[2]A. Bandura, D.H. SchunkCultivating competence, self-efficacy and intrinsic interest through proximal self-motivationJ. Abnorm. Soc. Psychol., 41 (1981), pp. 586-598View in ScopusGoogle Scholar[3]P. BennellLearning to change: Skills development among the economically vulnerable and socially excluded in developing countriesEmp. Train. Pap., 43 (1999)https://www.oit.org/wcmsp5/groups/public/-ed_emp/documents/publication/wcms_120372.pdfGoogle Scholar[4]N. Bennett, E. Dunne, C. CarréPatterns of core and generic skill provision in higher educationHigh. Edu., 37 (1) (1999), pp. 71-93, 10.1023/a:1003451727126View article View in ScopusGoogle Scholar[5]R. BennettEmployers’ Demands for Personal Transferable Skills in Graduates: a content analysis of 1000 job advertisements and an associated empirical studyJ. Vocation. Edu. Train., 54 (4) (2002), pp. 457-476, 10.1080/13636820200200209 View PDF View in ScopusGoogle Scholar[6]P. Brown, C. Lloyd, M. Souto-OteroThe Prospects for Skills and Employment in an Age of Digital Disruption: A CautionarySKOPE Publications, ISSN (2018), pp. 1466-1535https://files.eric.ed.gov/fulltext/ED593405.pdfGoogle Scholar[7]Chaitanya MallapurBudget 2019 Must Tackle Missed Skill Development Targets Even As Ministry’s Funds Go UnutilisedJanuary 29BloombergQuint; Bloomberg Quint (2019)https://www.bloombergquint.com/union-budget-2019/budget-2019-must-tackle-missed-skill-development-targets-even-as-ministrys-funds-go-unutilisedGoogle Scholar[8]Community Tool Box. (2020). Chapter 2. Other Models for Promoting Community Health and Development | Section 1. Developing a Logic Model or Theory of Change | Main Section | Community Tool Box. Ku.edu. https://ctb.ku.edu/en/table-of-contents/overview/models-for-community-health-and-development/logic-model-development/main.Google Scholar[9]L. Dacre Pool, P. SewellThe key to employability: developing a practical model of graduate employabilityEducation + Training, 49 (4) (2007), pp. 277-289, 10.1108/00400910710754435View article Google Scholar[10]A. DattaHow to Skill India: Youth don’t Want to Be welders, Their Aspiration is Drone Or Robot OperatorSeptember 10The Financial Express; The Financial Express (2017)https://www.financialexpress.com/education-2/how-to-skill-india-youth-dont-want-to-be-welders-their-aspiration-is-drone-or-robot-operator/849756/Google Scholar[11]Daugherty, L., Herman, R., & Unlu, F. (2017). Logic Models for Selecting, Designing, and Implementing Evidence-Based School Leadership Interventions Companion Guide to School Leadership Interventions Under the Every Student Succeeds Act. https://www.rand.org/content/dam/rand/pubs/tools/TL200/TL274/RAND_TL274.pdf.Google Scholar[12]Karina de Bruin, E. Cornelius, K BruinSelf-directed learning and career decision-makingActa Acad. Reeks B, 43 (2014), pp. 214-235ISSN 0587-2405Google Scholar[13]Demand-driven Training For Youth Employment toolkit Demand-driven Training For Youth Employment toolkit. (2017). https://youtheconomicopportunities.org/sites/default/files/contentupload/73868/Making_Cents_101717_Interactive_V3s.pdf.Google Scholar[14]Denney, A., Derner, S., Stump, S., & Inc, V. (2017). A Planning Guide for Aligning Career and Technical Education (CTE) and Apprenticeship Programs. https://s3.amazonaws.com/PCRN/reports/Planning_Guide_for_Aligning_CTE_and_Apprenticeship_Programs.pdf.Google Scholar[15]Dewan, S., & Khan, L. (2019). Breaking the Cycle of Vulnerability Education, Skills and Employability for Indian Youth. https://www.unicef.org/rosa/media/3926/file/Breaking%20the%20Cycle%20of%20Vulnerability.pdf.Google Scholar[16]G. Erabaddage, M. Shukri, A. KhatibiConceptualizing learning and employability “Learning and employability frameworkJ. Edu. Learn., 4 (2) (2015), 10.5539/jel.v4n2p532015(E-ISSN 1927-5269) View PDF Google Scholar[17]European Training Foundation. (2013). Work-based learning: benefits and obstacles a literature review for policy makers and social partners in ETF partner countries. https://www.etf.europa.eu/sites/default/files/m/576199725ED683BBC1257BE8005DCF99_Work-based%20learning_Literature%20review.pdf.Google Scholar[18]Ferns, S.J. (2018). Collaboration, cooperation and consultation: Work-integrated learning partnerships for enhancing graduate employability. https://doi.org/10.26182/5bbd68b462fb3.Google Scholar[19]M. Fugate, A.J. Kinicki, B.E. AshforthEmployability: A psycho-social construct, its dimensions, and applicationsJ. Vocat. Behav., 65 (1) (2004), pp. 14-38, 10.1016/j.jvb.2003.10.005View PDFView articleView in ScopusGoogle Scholar[20]D. GolemanAn EI-Based Theory of PerformanceC. Cherniss, D. Goleman (Eds.), The Emotionally Intelligent Workplace, Jossey-Bass (2001)ISBN-13: 978-0787956905http://www.eiconsortium.org/pdf/an_ei_based_theory_of_performance.pdfGoogle Scholar[21]A. Gupta, P. Mishra, C. Pandey, U. Singh, C. Sahu, A. KeshriDescriptive statistics and normality tests for statistical dataAnn. Card. Anaesth., 22 (1) (2019), p. 67, 10.4103/aca.aca_157_18View article Google Scholar[22]Hanita, M., Ansel, D., & Shakman, K. (2017). Matched-Comparison Group Design: An Evaluation Brief for Educational Stakeholders White Paper. https://www.edc.org/sites/default/files/uploads/matched_comparison_group_design.pdf.Google Scholar[23]Hargreaves, D.H. (2011). Leading a self improving school system. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/325890/leading-a-self-improving-school-system.pdf.Google Scholar[24]Hillage, J., & Pollard, E. (1998). Employability: Developing a framework for policy analysis. London: Department for Education and Employment DfEE. Research Report No. RR85.Google Scholar[25]Home | Ministry of Skill Development and Entrepreneurship | Government Of India. (2016). Msde.gov.in. https://msde.gov.in/en.Google Scholar[26]India Skills Report 2020. (2020). https://wheebox.com/assets/pdf/ISR_Report_2020.pdf.Google Scholar[27]International Labour Organisation, & World Bank. (2021). Skills Development in the Time of COVID-19: Taking stock of the Initial Responses in Technical and Vocational Education and Training United Nations Educational, Scientific and Cultural Organisation. https://www.ilo.org/wcmsp5/groups/public/-ed_emp/-ifp_skills/documents/publication/wcms_766557.pdf.Google Scholar[28]C.V. Izzo, J.P. Connell, M.A. Gambone, C.P BradshawUnderstanding and Improving Youth Development Initiatives through EvaluationS.F. Hamilton, M.A. Hamilton (Eds.), The Youth Development Handbook: Coming of Age in American Communities (2004), pp. 301-326, 10.4135/9781452232560.n13View article View in ScopusGoogle Scholar[29]S. Jung, S.J. Choi, J.C. Jeong, S.N. KimImpact of vocational education and training on adult skills and employment: An applied multilevel analysisInt. J. Educ. Dev., 66 (2019), pp. 129-138Chttps://econpapers.repec.org/article/eeeinjoed/v_3a66_3ay_3a2019_3ai_3ac_3ap_3a129-138.htmView in ScopusGoogle Scholar[30]S. KhuranaRedefining norms, exploring new avenues: Negotiations of women informal workers in DelhiInt. J. Comp. Sociol., 61 (2) (2020), Article 002071522094000, 10.1177/0020715220940009View article Google Scholar[31]D. Kneale, J. Thomas, K. HarrisDeveloping and Optimising the Use of Logic Models in Systematic Reviews: Exploring Practice and Good Practice in the Use of Programme Theory in ReviewsPLoS One, 10 (11) (2015), Article e0142187, 10.1371/journal.pone.0142187 View PDF View in ScopusGoogle Scholar[32]R. Kumar, S. Mandava, V.S. GopanapalliVocational training in India: determinants of participation and effect on wagesEmpir. Res. Vocat. Educ. Train., 11 (1) (2019), 10.1186/s40461-019-0078-y View PDF Google Scholar[33]E.S. KunnenThe Effects of Career Choice Guidance on Identity DevelopmentEdu. Res. Int., 2013 (2013), pp. 1-9, 10.1155/2013/901718Article ID 901718 View PDF Google Scholar[34]A. MangocheImpact of Technical and Vocational Education and Training (TVET) On Youth vulnerability in MalawiAcademia.edu (2015)https://www.academia.edu/10350609/Impact_of_Technical_and_Vocational_Education_and_Training_TVET_on_Youth_vulnerability_in_MalawiGoogle Scholar[35]R. Martin, F. Villeneuve-Smith, L. Marshall, E. MckenzieEmployability skills exploredwww.LSNeducation.org.uk (2008)http://fetn.org.uk/wp-content/uploads/2014/11/LSN-Employability-Skills-Explored.pdfGoogle Scholar[36]M.M. McClelland, F.J. Morrison, D.L. HolmesChildren at risk for early academic problems: the role of learning-related social skillsEarly Child. Res. Q., 15 (3) (2000), pp. 307-329, 10.1016/s0885-2006(00)00069-7View PDFView articleView in ScopusGoogle Scholar[37]Mehrotra, S., Raman, R., Kumra, N., Kalaiyarasan, D., & Röß. (2014). Vocational Education and Training Reform in India Business Needs in India and Lessons to be Learned from Germany. http://aei.pitt.edu/74136/1/Vocational_Education_and_Training_Reform_in_India.pdf.Google Scholar[38]A. Mitra, S. VerickYouth employment and unemployment: an Indian perspectiveInt. Lab. Org. (2013), Article 994806863402676https://EconPapers.repec.org/RePEc:ilo:ilowpsGoogle Scholar[39]J. O’Donoghue, T. MaguireThe individual learner,employability and the workplace—A reappraisal of relationships and propheciesJ. Eur. Indust. Train., 29 (9) (2005), pp. 436-446, 10.1108/03090590510610236View article View in ScopusGoogle Scholar[40]OECDThe Impact of COVID-19 On Student Equity and inclusion: Supporting vulnerable Students During School Closures and School Re-OpeningsOECD (2020)https://www.oecd.org/coronavirus/policy-responses/the-impact-of-covid-19-on-student-equity-and-inclusion-supporting-vulnerable-students-during-school-closures-and-school-re-openings-d593b5c8/Google Scholar[41]PCRN: Employability Skills. (2021). Ed.gov. https://cte.ed.gov/initiatives/employability-skills-framework.Google Scholar[42]M.M. Pheko, K. MolefheAddressing employability challenges: a framework for improving the employability of graduates in BotswanaInt. J. Adolesc. Youth, 22 (4) (2016), pp. 455-469, 10.1080/02673843.2016.1234401 View PDF Google Scholar[43]D.H. Schunk, F. PajaresSelf-Efficacy Beliefs in International Encyclopedia of Education(Third Edition) (2010), pp. 668-672, 10.1016/b978-0-08-044894-7.00620-5View PDFView articleView in ScopusGoogle Scholar[44]Skills Development. (2021). World Bank. https://www.worldbank.org/en/topic/skillsdevelopment.Google Scholar[45]The impact of COVID-19 on student equity and inclusion: Supporting vulnerable students during school closures and school re-openings. (2020). OECD. https://www.oecd.org/coronavirus/policy-responses/the-impact-of-covid-19-on-student-equity-and-inclusion-supporting-vulnerable-students-during-school-closures-and-school-re-openings-d593b5c8/.Google Scholar[46]UNESCO. (2018). Quick Guide to Education Indicators for SDG 4. http://uis.unesco.org/sites/default/files/documents/quick-guide-education-indicators-sdg4-2018-en.pdf.Google Scholar[47]W.K.Kellogg Foundation Logic Model Development Guide. (2004). https://www.aacu.org/sites/default/files/LogicModel.pdf.Google Scholar[48]M. Yorke, P.T. KnightEmbedding Employability Into the curriculum, Learning and Employability, Series 1The Higher Education Academy (2006)ISBN: 1-905788-00-2https://www.qualityresearchinternational.com/esecttools/esectpubs/yorkeknightembedding.pdf Google Scholar