From College-to-Work: Latent Class Models Analysis of Mutual Adjustment in Internships after the Diploma

Abstract

This article aims to identify which factors enable the mutual adjustment of top managers of small enterprises (SEs), frequently their owners, and interns (young college graduates) in their first work experiences. Using latent segment models and data obtained from an archive database of an internship program with complete information on both interns and SEs top managers (n = 102), the categories of mutual adjustment between interns and SEs top managers were obtained: (1) mutual adjustment corresponding to very satisfied SE top managers and interns, hired at the end of the internship (75%); and (2) mutual maladjustment characterized by the dissatisfied SE top managers and interns, resulting in no job creation (25%). Finally, we conducted a binary logistic regression (mutual adjustment as the dependent variable), using forward LR, and we identified organizational socialization practices and learning system as factors that contribute to explain mutual adjustment.

1. Introduction

School-to-work transition has been an important subject of research in Sociology of Work and Sociology of Youth. In these domains, the use of surveys of young people enables us to trace their transition paths from school to work. However, these studies do not solve the “black box” of how the first work experiences influence youth employment. Focusing on organizational socialization enables us to rethink contextually the youth employment problems. Van Maanen and Edgar Schein [1] introduced the concept but the assumptions they provided have never been fully examined. Data collection has been one the major limitations of empirical research on organizational socialization. Furthermore, the phenomenon is frequently studied only through newcomers’ self-reports [2,3,4,5]. Following an inductive approach, as Bauer and Taylor [6] suggest, this article presents a framework of analysis combining the perceptions of both newcomers and top managers. As a result, neither an under-socialized nor an over-socialized perspective is emphasized [7,8]. In addition, considering the mutual adjustment concept as the result of the socialization process allows us to diminish the ambiguity in distinguishing the two concepts. The term socialization can therefore be used to refer exclusively to the process and not to the result to be achieved [9,10,11].

Although theoretically defined [12], mutual adjustment has not been empirically tested. In fact, recent studies have been published about newcomers’ adjustment with an innovating analysis focus [13,14,15], but none of the studies adopt the concept of mutual adjustment, questioning newcomers only [16]. Our study extends on previous work in this way and bearing in mind the literature review on organizational socialization and Wanous’ model [12], in this article mutual adjustment is reciprocal and has two analytical dimensions: (1) SE top manager and (2) Newcomer.

The aim of this paper is to identify which factors enable the mutual adjustment of top managers of small enterprises (SEs), frequently their owners, and interns (young college graduates) in their first work experiences. These first experiences affect youth employment and reflect youth difficulties in adapting and the shock they feel with the reality of work [17,18]. For this reason, organizational socialization assumes particular importance for newcomers, even after 18 months [19]. More recent studies present findings that go in the same direction [20,21].

Considering the relevance and implications of the topic for educational policies and human resources management, it is surprising that more recent research is scarce. Though recent studies can be cited [5,20,21,22,23,24], it remains that one of the major limitations of previous empirical studies on organizational socialization and adjustment is the exclusive use of newcomers’ self-reports, even in recent publications [5]. Our research thus extends the literature and knowledge on mutual adjustment, considering employers and interns’ perceptions.

There is no consensus on whether or not internships result in employment. During the 2008 financial crisis, the worsening of youth unemployment justified the adoption of many internship measures, to prevent young people from becoming long-term unemployed or discouraged, thus reducing their career prospects and potential salaries [25]. However, it should be noted that these internships were not aimed at job creation. With the COVID-19 pandemic, many internships were canceled or changed in the way they were carried out. In spite of this, it is possible today to have greater knowledge about the effects of internships on the employability of young people [26]. For example, data is now available on the short-term negative effects of the quality of employment after the internship, both in terms of stability and in terms of salary levels [27,28] or labor protection [29]. This article, focusing on interns’ mutual adjustment in small and medium-sized enterprises, can however, bring other insights into a better understanding of which factors determine the chances of internships resulting in jobs.

2. Literature Review

Wanous [12] conceives that organizational adjustment depends primarily on the appraisal of workers’ performance, affected by job design and workers’ skills fit. On the other hand, newcomers’ adjustment depends on job satisfaction which, in turn, will influence organizational commitment. Moreover, job satisfaction and organizational commitment, separately or collectively, influence interns’ decisions of remaining or leaving the organization.

The mutual adjustment concept, combining the perceptions of the system and agents, rejects the deterministic perspectives of organizational socialization and establishes the view of reciprocity in labor relations. More than a formal contract, labor relations involve a tacit understanding between workers and employers, a psychological contract [30], which influences the behavior of both. Thus, organizational socialization of inclusion may be influenced by that psychological contract. The intentions of both newcomers and organizations in establishing labor relations can also be influenced by this early stage of socialization [31,32]. Other empirical studies have analyzed the effects of socialization on person-organization fit, but do not define this concept [33]. Saks and Ashforth [34] examined the influence of individual and situational factors on newcomers’ adjustment. Although they did not define adjustment, the variables they used give an approximate idea of the concept. Job satisfaction, organizational commitment, organizational identification, intentions to quit, frustration, stress symptoms and job performance are used to assess adjustment [34]. Results suggested that situation is the primary determinant of newcomers’ adjustment to work [34]. In a previous study by these authors, adjustment was assessed through job satisfaction, intentions to quit (or the reverse), organizational commitment and organizational identification [3]. Some of these variables have been widely used to assess newcomers’ adjustment. Feldman [35] distinguishes behavioral from affective outcomes. The behavioral outcomes include (a) carry out role assignments; (b) remain with the organization, (c) innovate and cooperate spontaneously to achieve organizational objectives that go beyond role specifications; and the affective outcomes comprise (a) general satisfaction, (b) internal work motivation, and (c) job involvement. Cooper-Thomas and Anderson [18] consider job satisfaction, organizational commitment and intention to quit. For Adkins [17], the outcomes of the socialization process are job satisfaction, organizational commitment and turnover.

Literature review has revealed that there is no consensus on the influence of individual factors on adjustment. Some studies confirm the influence of those factors on organizational commitment [36]; others highlight the moderating effect of training and guidance on newcomers’ anxiety [37]. However, other studies emphasize the influence of situational factors on organizational adjustment [3], on performance, organizational satisfaction and commitment [38], though some individual variables can also affect adjustment to the organization. All in all, individual factors considered in empirical research are extensive, making it difficult to draw conclusions.

According to Jones [39], organizational socialization is to be measured by the intentions of the newcomers and the socialization agents. Later, in an empirical study, the author suggested that the social dimensions of socialization, as investiture and serial processes, influenced the adjustments to organizations. The study also found that individual factors moderated the effects of institutionalized tactics [40]. The influence of the socialization content has also been considered as a moderator of organizational socialization tactics and newcomers’ adjustment. Task mastery, role clarity, and group integration, seem to be related to organizational commitment and work quitting [41]. Cooper-Thomas and Anderson [18] found that the socialization content—information acquisition—mediated organizational socialization tactics with both job satisfaction and organizational commitment. These authors thus proposed that adjustment should be studied considering also a content approach [42].

Some studies suggest three modes of adjustment: person-job, person-group and person-organization, considering three levels of analysis [43,44]. However, this construct does not assume adjustment as a reciprocal phenomenon. In turn, Black and Ashford [45] identified two modes of adjustment: self-change and job-change, taking into account reciprocal adjustment already presented by Schein [30] and Nicholson [46].

3. Materials and Methods

The objective of the study is to identify which factors enable the mutual adjustment of top managers of small enterprises (SEs), and interns in their first work experiences. This objective has an underlying concern to know what influences job creation in small companies through internships?

According to Eller, Gerber & Robinson [47] the research purpose is to produce a clear and meaningful understanding and explanation of social phenomena. Furthermore, the analysis of cases that present rare characteristics can contribute to a better understanding of phenomena [48]. The interest in researching this program stands from the additional supervision provided by the program consultants who periodically examined the workplaces, thus increasing the quality of internships [49] and potentially generating new youth employments. The program consisted in providing support to small and medium-size companies in their placement of young recent graduates in nine-month internships.

3.1. Procedures

As the program was discontinued a few years ago, the authors of the present paper requested access to the archived program documentation. All documentation regarding 2006, the penultimate year in which the program operated, was made available for research purposes. That year, the internship program supported 236 internships on small enterprises (up to 49 employees). Each enterprise had only one intern and the SE top manager was frequently the internship supervisor. All internships with complete information on intern and SE top manager were considered: 102 internships (43% of the overall internships provided by the program).

The documentation provided included internship reports by the program consultants, interns and entrepreneurs’ C.V.’s, application forms and monitoring reports, and interns and entrepreneurs’ satisfaction surveys. An inductive approach was thus followed in this study. Several empirical studies recommend inductive approaches to the study of organizational socialization [4,6]. According to Lee [50] several reasons may justify the use of unobtrusive methods: (1) direct data collection, sometimes confidential, can raise doubts of non-confidentiality and lack of anonymity, affecting reliability of the data; on the other hand, (2) it allows access to data, not biased for research purposes. Written consent was obtained to access the information for research purposes.

Throughout an interactive process of literature review and pre-analysis of available documentation relevant data for analysis was selected which thus stimulated the review of literature not originally planned. Moreover, the procedures recommended for this method were followed [50,51], thus establishing an open contact with the former program managers and former program consultants. Exploratory interviews were also conducted with former interns and entrepreneurs that participated in the program. Doubts about documents and data contained therein were therefore explained, namely concerning the origin of the documents, their authors and reliability, when and how they were produced, and for what purpose.

Data analysis proceeded first to achieve latent variables. For that clustering methods by estimating latent class models (LCM) were used because they present several advantages over traditional methods [52,53], through software Latent Gold 5.0. Next, with all the necessary variables, including the latent variable of mutual adjustment (our framework of analysis’ dependent variable), in order to identify which factors contribute to justify the mutual adjustment, binary logistic regression using SPSS 28.0 software and Forward-LR as selection method were applied.

Our analysis focused on 102 small enterprises, of which 55% had up to 9 employees, and 45% had from 10 to 49 employees; 57% were family businesses and 43% non-family businesses; 39% were in the secondary sector and 57% in the tertiary sector; almost all (92%) were private companies limited by shares; and in terms of existence 29% were less than 5 years old, 23% were 5 to 8 years old, 27% were 8 to 17 years old and 21% had existed for over 17 years.

The interns, 52% women and 48% men, were, on average, 26 years old, almost all single (97%) and 88% had less than one year of professional experience. These interns came from different academic fields (42% were from Social Sciences, Commerce and Law and 35% from Engineering, Manufacturing Industries and Building) and 89% studied in the public sector of tertiary education. Seventy percent of the SE top managers, who supervised the internships, were men and they were, on average, 40 years old. Fifty percent had tertiary education and the large majority (78%) assumed they were the SEs owners.

3.2. Measures

Mutual adjustment comprises two analytical dimensions: the SEs top managers’ and the interns’ perceptions about the internship. In the organization’s analytical dimension two variables are included: (a) SEs top managers’ rating on appraisal of interns’ performance, a five-point Likert scale; and (b) SEs top managers’ hiring intentions, a dichotomous variable. In our opinion, SEs top managers’ satisfaction with interns’ performance can be measured by workers’ performance appraisal as considered by Wanous [12]. However, this variable may comprise technical and social skills’ assessment.

The interns’ analytical dimension consists of: (a) job satisfaction and (b) interns’ hiring intentions. Job satisfaction is a latent variable from several observed variables: (a) Job enlargement interest; (b) SE top manager or entrepreneur support and/or interpersonal relations; (c) Senior staff support and/or interpersonal relations; (d) Workers’ support and/or interpersonal relations; (e) Program consultant/tutor support and/or interpersonal relations.

As information on SE top managers’ and interns’ hiring intentions was available, it was possible to determine which internship factors contributed to youth employment. The hiring intentions of both SE top managers and interns are dichotomous variables. Those hiring intentions must not be thought of just in terms of the law, but also as the establishment of a psychological contract. Therefore, flexible working forms may result from internships. Whenever internships do not result in a new job, both SE top managers’ and interns’ intentions and reasons should be studied.

As individual factors, variables such as age, sex and level of education of both SEs top managers and interns were included. Tertiary education sector of the interns was also included. Some studies suggest that perceptions of adjustment can be influenced by the number of job offers [33]. Therefore, in this model, more attractive job offers are considered to be a possible influence on interns’ intentions to quit the organization. Those job offers were an open question on the interns’ survey and were later classified through a content analysis.

As situational factors, the demographic attributes of organizations, such as number of employees, sector, legal nature, and social capital were included. The size and nature of the organization were suggested by Jones [40] as to influence socialization tactics. Ashforth, Saks and Lee [3] studied the effects of some of these variables. In this article, the role of SEs top managers in the supervision of internships was also considered, with three dichotomous variables: “Was the SE top manager the internship supervisor?”; “Was the SE top manager enrolled in the intern selection process?” and “Was the SE top manager enrolled in the welcoming/integration in enterprise?”. The SE top managers’ initial objectives for participating in the program were also included in the analysis, as dispositional factors. The SE top managers’ initial objectives were measured by 17 dichotomous variables they answered in one of the application forms to the program. The aim was to determine if those initial objectives influence mutual adjustment.

Organizational variables reflecting economic and financial conditions were provided by the reports of the program’s consultants. The aim was to determine if economic and financial factors influenced youth employment in SEs. The consultants included a qualitative and a quantitative analysis of the organizational problems in their diagnostic reports. In this study, their quantitative assessments of the enterprise were included for analysis. There were 8 items in a 4-point scale.

The internship program included formal and collective organizational socialization tactics, through formal training sessions, and informal and individual tactics, in the workplace. Since the socialization practices reflect interns’ perceptions, they are not considered as contextual factors, as in other studies [2,40,54]. The socialization practices were measured by 3 items in a 4-point scale, from the interns’ survey, in order to assess the interns’ perceptions of the different kinds of training actions: the welcoming/integration in enterprise, the in-class training and also the on-job training.

We can thus formulate the following hypotheses:

Hypothesis 1.

There is heterogeneity in Organizational Socialization Practices according to the following indicators: Welcoming/integration in enterprise, In-class training and On-job training.

Collective and individual tactics have been studied in a mutually exclusive way [1,5,18,23,40]. However, this internship program combined these two techniques, In-class training as a collective tactic and On-job training as an individual tactic, providing an opportunity to analyse their effects on socialization.

Hypothesis 2.

There is heterogeneity in Interns’ Job Satisfaction, according to the following indicators: Job enlargement interest, SE top manager or entrepreneur support and/or interpersonal relations, Senior staff support and/or interpersonal relations, Workers support and/or interpersonal relations, and Program consultant/tutor support and/or interpersonal relations. Research has shown that the support of workers [19,23] and supervisors [55] influences the adjustment of new hires, so it is important to consider these factors.

Hypothesis 3.

There is heterogeneity in Mutual Adjustment according to the following indicators: Interns’ job satisfaction, Hiring and SE top managers’ rating on intern performance appraisal. This hypothesis is based on on Wanous’ [12] organizational mutual adjustment model and also on person-organizational fit literature.

Hypothesis 4.

Organizational socialization practices directly influences Mutual Adjustment. This hypothesis derives from empirical studies that have demonstrated the effects of socialization on person-organization fit [5,33,34].

Hypothesis 5.

Tertiary education sector directly influences Mutual Adjustment.

The formulation of Hypothesis 5 is justified because different sectors may be associated with different institutional habitus [56]. In fact, Çelik [56] considers that schools can influence students’ educational trajectories. Considering the influence of tertiary education sector on mutual adjustment, this study extends previous research. It is assumed that different sectors may be linked to different institutional habits.

4. Results

4.1. Organizational Socialization Variable

The variable “practices of socialization and integration” is a latent variable, which integrates indicators referring to welcoming and integration in the enterprise, to in-class training received with other interns, involving formal and collective socialization tactics, using the classification of Van Maanen and Schein [1], and to in-work training, corresponding to those authors’ informal and individual socialization tactics (see

Table 1. Organizational Socialization Practices—Estimated Parameters.

Variables	Indicators	Cluster 1 (61%)	Cluster 2 (39%)
Welcoming/integration in enterprise	Poor	0.0134	0.1987
	Fair	0.1483	0.3972
	Very good	0.8383	0.4041
In-class training	Unadjusted	0.0033	0.0223
	Insufficient	0.0541	0.1617
	Fair	0.3807	0.4965
	Very good	0.5620	0.3194
On-job training	Unadjusted	0.0006	0.1090
	Insufficient	0.0383	0.4340
	Fair	0.6565	0.4446
	Very good	0.3047	0.0123

).

This variable will later be used in the analysis model, as independent variable in the identification of the factors which influence mutual adjustment. Through the estimation of latent class models-one-cluster latent model or homogeneity model, and models from this one to four-cluster latent model—the information criteria AIC3 [53] selected the model with two clusters.

Table 1 shows the parameter estimates of the selected model (probabilities of belonging to clusters, 0.61 and 0.39, respectively, and the probabilities of answering to each category of variables, given that it belongs to a cluster).

The conditional probabilities must be interpreted as follows: in the welcoming/integration in the enterprise variable, 0.8383 and 0.4041 represent probabilities of answering very good within cluster 1 and 2, respectively. As 0.8383 is the largest value of those probabilities, it is concluded that very good is a characteristic of cluster 1 (

Table 2. Interns’ Satisfaction Profiles with Organizational Socialization Practices.

Variables	Cluster 1 Very Satisfied	Cluster 2 Dissatisfied
Welcoming/integration in the enterprise	Very good	Poor/Fair
In-class training	Very good	Unadjusted/Insufficient/Fair
On-job training	Fair/Very good	Unadjusted/Insufficient

).

The latent class model permitted the identification of two profiles, one encompassing the more favorable assessments, 61 percent of interns, and the other corresponding to the dissatisfied assessments, roughly 39 percent of interns.

The indicators which were used, in view of the estimation of the conditional probabilities, allowed the classification of the given profiles, in this case very satisfied, 61 percent and dissatisfied, 39 percent (see Table 2).

The individuals classified as dissatisfied considered the welcoming and integration as poor or fair—roughly 60 percent of individuals in this profile, in-class training was assessed as unadjusted, insufficient or fair, roughly 70 percent, and on-job training as unadjusted or insufficient, roughly 54 percent of individuals in this profile.

This result allows us to accept H₁, thus concluding that there is heterogeneity in the practices of socialization and integration, according to the following indicators: Welcoming/integration in enterprise, In-class training and On-job training.

4.2. Interns’ Job Satisfaction Variable

Interns’ job satisfaction is a latent variable estimated from interns’ job enlargement interest, and interns’ perceived support and/or interpersonal relations with SE top manager or entrepreneur, senior staff, workers and program consultant/tutor. The estimation of latent class models permitted the attainment of the categories of that variable and tests the hypothesis of heterogeneity of interns on satisfaction. Three interns’ job satisfaction profiles, with 12, 44 and 44 percent of the interns, respectively (see

Table 3. Interns’ Job Satisfaction—Estimated Parameters.

Variables	Indicators	Cluster 1 (12%)	Cluster 2 (44%)	Cluster 3 (44%)
Job enlargement interest	Not at all interesting	0.003	0.0524	0.0017
	Slightly interesting	0.0157	0.0949	0.0105
	Interesting	0.1722	0.3608	0.1403
	Very interesting	0.604	0.44	0.5996
	Extremely interesting	0.205	0.0519	0.2479
SE top manager or entrepreneur support and/or interpersonal relations	Hostile	0.9388	0.1486	0
	Poor	0.0555	0.2038	0.0004
	Fair	0.0056	0.4741	0.0461
	Welcoming	0.0001	0.1736	0.9535
Senior staff support and/or interpersonal relations	Hostile	0.7452	0.0107	0
	Poor	0.237	0.153	0.0002
	Fair	0.0175	0.5115	0.0328
	Welcoming	0.0002	0.3248	0.967
Workers support and/or interpersonal relations	Hostile	0.7256	0.0161	0.0001
	Poor	0.0981	0.0262	0.0013
	Fair	0.1654	0.5319	0.1633
	Welcoming	0.011	0.4259	0.8353
Program consultant/tutor support and/or interpersonal relations	Hostile	0.8536	0.0854	0.0048
	Poor	0.1295	0.2005	0.0378
	Fair	0.016	0.3832	0.2442
	Welcoming	0.0009	0.3309	0.7133

) were obtained. The null hypothesis of homogeneity was rejected, thus concluding in favor of heterogeneity.

The very dissatisfied with 12 percent, the moderately satisfied with 44 percent and the very satisfied with 44 percent, as shown in Table 3 were thus identified.

Although interns of cluster 1 (very dissatisfied) are shown to be satisfied with job enlargement interest, they are extremely dissatisfied with support and/or interpersonal relations (particularly with the SE top manager and with the program consultant/tutor, i.e., those responsible for supervising the internship).

Interns classified in the group of satisfied (cluster 2) with internship were the most dissatisfied with job enlargement interest. However, they have proven to be satisfied with the social relations established during the internship. Note that, as to social relations, the most frequent answer is fair. Additionally, in this profile, relations with the SE top manager/entrepreneur and the program consultant/tutor were less satisfactory.

The very satisfied interns (cluster 3) expressed high levels of interest with job enlargement (see

Table 4. Interns’ Job Satisfaction Profiles.

Variables	Cluster 1 Very Dissatisfied	Cluster 2 Satisfied	Cluster 3 Very Satisfied
Job enlargement interest	Very interesting	Not at all interesting/Slightly interesting/Interesting	Extremely interesting
SE top manager or entrepreneur support and/or interpersonal relations	Hostile	Poor/Fair	Welcoming
Senior staff support and/or interpersonal relations	Hostile/Poor	Fair	Welcoming
Workers support and/or interpersonal relations	Hostile/Poor	Fair	Welcoming
Program consultant/tutor support and/or interpersonal relations	Hostile	Poor/Fair	Welcoming

). About 60 percent of individuals in this profile were found to have great interest in their work, and 25 percent revealed even more interest in the functions performed.

All the variables used in order to obtain the latent variable Interns’ Job Satisfaction Profiles will necessarily be related to it. To find out which of them were more correlated to Interns’ Job Satisfaction Profiles, the Spearman’s coefficient correlations were calculated. Although all the indicators that compose the latent variable interns’ job satisfaction influence it significantly, the interpersonal relations are the ones which most influence interns’ job satisfaction, particularly those relating to SE top Manager (0.887) and Senior Staff (0.820).

Again, these results allow the acceptance of H₂, proving the existence of heterogeneity in the Interns’ Job Satisfaction according to the following indicators: Job enlargement interest, SE top manager or entrepreneur support and/or interpersonal relations, Senior staff support and/or interpersonal relations, Workers support and/or interpersonal relations, and Program consultant/tutor support and/or interpersonal relations.

4.3. Mutual Adjustment Variable

Mutual adjustment between interns and SE top managers is the dependent variable. The indicators make use of both perspectives, the interns and also SE top managers. Being a latent variable to obtain, based on those indicators, the technique of clustering is used again, via estimation of latent class models [52]. The optimal solution, by using information criterion AIC3 [53] for model selection, is a model with two- latent clusters, with the parameters’ estimates displayed in

Table 5. Mutual Adjustment—Estimated Parameters.

Variables	Indicators	Cluster 1 (75%)	Cluster 2 (25%)
Interns’ job satisfaction	Very dissatisfied	0.0511	0.2868
	Satisfied	0.4195	0.5507
	Very satisfied	0.5294	0.1626
Hiring	Yes	0.9098	0.0587
	No	0.0902	0.9413
SE top managers’ rating on intern performance appraisal	Poor	0.0012	0.3828
	Fair	0.0168	0.2804
	Good	0.3538	0.3082
	Very good	0.6283	0.0286

.

Seventy-five percent of the cases were classified in cluster 1, which includes internships that provided youth employment, whereas the remaining are in cluster 2, including internships without a contract at the end of the internship (Table 5).

It is observed (

Table 6. Profile of Mutual Adjustment.

Variables	Cluster 1 Mutually Adjusted	Cluster 2 Mutually Unadjusted
Interns’ job satisfaction	Very satisfied	Very dissatisfied/Satisfied
SE top managers’ rating on intern performance appraisal	Good/Very good	Poor/Fair
Hiring	Yes	No

) that cluster 1, which have been named the mutually adjusted (75%), includes the internships in which the interns were very satisfied and the SE top managers classified the interns’ performance as good or very good resulting in employment.

On the contrary, profile 2 was classified as the mutually unadjusted (25%), involving the internships in which the interns were very dissatisfied or satisfied with the internship and SE top managers’ rating on performance appraisal varied between poor and fair, without hiring at the end of the internship.

Finally, the concept of the mutual adjustment emphasizes the congruence of SE top managers’ and interns’ perceptions. The profile of mutually adjusted internships corresponds to the SEs top managers and interns who were more satisfied with the experience, which led to the hiring of the intern, and the profile of mutually unadjusted corresponds to the SEs top managers and interns who least appreciated the experience, with no hiring outcome.

It can thus be concluded that there is heterogeneity in Mutual Adjustment (H₃), according to the following indicators: Interns’ job satisfaction, Hiring and SE top managers’ rating on intern performance appraisal.

4.4. Mutual Adjustment: What’s Inside the Box?

After trying to understand the factors associated with each of these indicators, it is relevant to analyze those which most contribute to explain mutual adjustment.

A Binary Logistic Regression was conducted, because the dependent variable (mutual adjustment) is binary, using forward LR, integrating dispositional, contextual and individual factors of intern and of the SE top manager with the role of the tutor. In order to identify which factors enable the mutual adjustment of top managers of small enterprises (SEs), frequently their owners, and young college graduates in their first work experiences, affecting youth employment, factors such as the following were included: (1) aspects of organizational socialization of interns and/or with existing problems of the SE where the internship took place, its nature (family/non-family business), number of employees and age-contextual factors, (2) SE top managers’ attitudes profiles-dispositional factor, (3) SE top managers’ age and qualifications—SE top managers’ individual factors, (4) interns’ age, gender, academic degree, tertiary education sector, time after graduation, interns’ previous professional experience, duration of internship– intern’s individual factors.

The values of Hosmer and Lemeshow test (p-value = 0,82), Cox & Snell R Square = 33.6% and Nagelkerke R Square = 49% show that it is a good model.

From

Table 7. Variables in Equation—Mutual Adjustment.

Variables in Model	β	D.P.	Wald	GL	p-Value	Exp(β)
Organizational socialization practices (1)	3.622	1.158	9.779	1	0.002	37.426
Tertiary education sector			3.751	2	0.153
Private tertiary education sector	1.524	2.120	0.517	1	0.472	4.591
Catholic University of Portugal	3.335	1.768	3.560	1	0.050	28.087
Constant	−3.335	1.060	9.894	1	0.002	0.036

it is possible to observe that organizational socialization practices, and tertiary education sector, namely private tertiary education and Catholic University of Portugal sector categories were selected for the model, through the Forward LR method, thus being important factors to explain mutual adjustment. As such, it is concluded that organizational socialization practices (p-value = 0.002) and tertiary education sector, namely Catholic University of Portugal category (p-value = 0.05) help to explain mutual adjustment, sustaining H₄ and H₅. These results suggest that the situational factors are the most influential on internships. Moreover, in relation to the tertiary education sector, also selected through the method Forward-LR (p-value = 0.050), it would be advisable to include it in further studies to best determine its influence, since, according to this model, the Catholic University of Portugal category influences mutual adjustment more than the categories Public and Private.

Thus, given the selected indicators, it is not possible to identify predispositions of SE top managers favoring the subsequent mutual adjustment. Interns’ and SE top managers’ individual factors such as age, qualifications, and others were also not relevant to explain the dependent variable. Furthermore, the characteristics of the SEs were also irrelevant in explaining mutual adjustment, suggesting that it occurs regardless of their size, age, nature (family/non-family business), and management problems identified by the program consultants.

The results obtained through Binary Logistic Regression motivated the calculation of the correlations between these variables and mutual adjustment, including also career prospects (

Table 8. Spearman’s Coefficients’ Correlations Interns’ career prospects and other variables.

Spearman Correlation p-Value (Two-Sided)	(1)	(2)	(3)	(4)
		0.108	0.223 *	0.314 **
Interns’ career prospects in SEs (1)	1	0.352	0.031	0.003
Mutual Adjustment (2)		1	0.348 **	0.429 **
			0.001	0.000
Tertiary education sector (3)			1	0.156
				0.138
Organizational socialization practices (4)				1

*. Correlation is significant at the 0.05 level; **. Correlation is significant at the 0.001 level.

). Indeed, mutual adjustment is significantly correlated with both tertiary education sector (0.378) and organizational socialization practices (0.429). Tertiary education sector is also significantly correlated with interns’ career prospects in SEs (0.223), which indicates that some tertiary education sectors may prepare their students differently for school-to-work transition. According to the Binary Logistic Regression results, students from the Catholic University of Portugal are more sensitive to working in small enterprises (up to 49 employees).

5. Discussion

Informal and individual on-job training and the period of welcoming/integration in the organization were the most influential variables of the profiles of organizational socialization practices (H₁ sustained). The internship program provided both collective and individual on-job-training, making it possible to determine which is more influential on organizational socialization. The findings are consistent with those of Jones [40] which emerged as a particularly significant influence of investiture and serial socialization tactics. However, Cable and Parsons [33] found out that contextual tactics (collective and formal versus individual and informal) were unrelated to interns’ subjective Person-Organization fit perceptions. The lack of congruence of the findings may result from methodological differences. More recently, Kowtha [5] stated that many studies have focused on only one of the tactics which obfuscates the organization’s socialization tactics implications. However, focusing on how the socialization tactics are put into practice could also provide a more comprehensive awareness of such diversity in the results. For example, spontaneous training sessions and unqualified tutors are crucial to workplace training [57,58,59,60]. This focal point of research is even more relevant in SEs where training tends to occur on job [61]. The significance of on-job-training is also relevant because of the regular supervision the program provided to interns.

Interns’ Job Satisfaction also emerged as a heterogeneous variable (H₂) and all the indicators were relevant for the emergence of the three profiles of that latent variable. However, the interpersonal relations are the ones which most influence interns’ job satisfaction, particularly those relating to SE top Manager and Senior Staff. These results are consistent with those found by other studies [20,23,55], even though those studies did not focus on small businesses. It should be kept in mind that in small companies it is less likely that there are senior managers who can take on the role of mentors, coaches, or even role models to facilitate the adjustment of interns. It is therefore especially difficult to integrate young executives in small companies. As such, universities can consider preparing their students for these difficulties, and also adjust their employment expectations, given that in Portugal most private companies are small.

Mutual Adjustment is a heterogeneous variable (H₃) emerging two clusters, according to interns’ job satisfaction, and SE top managers’ rating on intern performance appraisal, and the hiring decision of both interns and SE top managers. This concept shapes a view of reciprocity in the work relationships established within professional (graduate) internships and helps to break an important limitation of studies on organizational socialization and newcomer adjustment that are limited to the analysis of interns’ self-reports [2,3,4,5]. The concept of mutual adjustment allows us to see that the (in)satisfaction of trainees and the entrepreneurs are related to the intention of both to continue or not to continue an employment relationship after the end of the internship. We also found that the variables that contribute most to the discrimination of profiles are hiring at the end of the internship and the manager’s evaluation of the intern’s performance. This result is also pertinent because if the entrepreneur does not want to hire, even if the trainee considers staying on with the company, there is no job creation. That allows us to recall Berger e Luckman expression ‘he who has the bigger stick has the better chance of imposing his definitions of reality’, cited by Wentworth [10]. This finding means that the newcomer role definition depends primarily on the SE top manager, who has the final decision, even if the newcomers try to influence the decision.

Organizational socialization practices emerged as the main factor that clearly influences the mutual adjustment profiles (H₄ sustained), followed by tertiary education sector (H₅ sustained). The significant impact of organizational socialization practices on mutual adjustment, in the present findings, stresses the relevance of the organizational entry stage to the turnover, as Adkins [17], and more recent studies concludes [23,32]. It suggests the need to carefully plan internships, namely the organizational entry period and also the on-job training planning and guidance. Some internship features can be decisive to create new jobs in SE at the end of that training period. They can be described as, properly setting down the objectives of internships; preparing coaching and tutoring actions, having external tutors which are also useful to mediate possible conflicts between interns and SE top managers; giving previous training to SE top managers in guiding and coaching interns; also previously giving interns preparation on what an internship is and how to work in a SE. These on-job structured training sessions may solve several problems experienced by interns and SE top managers.

Young college graduates’ unemployment rates have been rising over time and, to cope with it, several internship programs have been recently launched in Portugal. However, these programs are mainly designed to put young people in large companies or in the Public Administration, where the capacity to absorb the interns is restricted. In our view, internship programs should be focused on the creation of youth employment in SEs. Additionally, teachers and other educational agents should prepare college students to work in SEs, making them aware of career prospects in small-sized enterprises, for example through classroom exercises, case studies and other teaching methods.

Regarding the variable tertiary education sector, also selected by the method Forward-LR (p-value = 0.059), it is recommended that it be included in subsequent studies to better determine their influence since, according to the Binary Logistic Regression model, Catholic University education category influences mutual adjustment more than public and private education categories. These results justified the correlations between some variables and mutual adjustment were calculated, also including prospects of a continuing career, in a methods triangulation strategy. Indeed, mutual adjustment is significantly correlated with both tertiary education sector and organizational socialization practices; tertiary education sector is also significantly correlated with interns’ career prospects in SEs. This correlation indicates that some tertiary education sectors may prepare their students differently for school-to-work transition, which is consistent with the idea that institutional habitus may influence students in their employment trajectories. The findings, according to the Binary Logistic Regression results show that the students from Catholic University of Portugal are more sensitive to working in small enterprises. It might be interesting to better understand how this university prepares its students in this direction.

One of the major limitations of early empirical studies on organizational socialization is the use of interns’ self-reports [16]. This empirical study, aiming to identify which factors encourage or hinder youth employment in those organizational contexts, shows that the analysis of the perceptions of both “employees” and “employers” is also relevant in terms of employment [62,63]. Therefore, future studies should include both perspectives to gain deeper comprehension on internships job creation. Additionally, in small enterprises, decision-making tends to be centered on the SE top manager, who is frequently the owner. These findings can be useful for the purpose of youth employment policies, particularly when internships are involved, and also for internship evaluation purposes, considering not just interns’ self-reports.

In view of the impact of social interactions in mutual adjustment, developing emotional and social skills seems to be relevant. According to Louis [64], the “surprise” felt by interns entering non-familiar organizational contexts is cognitive but also emotional, since their experiences differ from insiders’. For the young, this social learning is particularly difficult as they have been prolonging their education and postponing their school-to-work transition [65]. Moreover, as newcomers’ interactions with supervisors and co-workers are centered on role negotiation [66], their emotional skills should be emphasized and developed.

The use of unobtrusive methods implies accepting the inherent limitations and this motivated the careful selection of documents which were subject to internal reliability tests. However, these methods allowed the simultaneous study of the opinions of interns and SE top managers without compromising anonymity perceptions and consequently data reliability.

6. Conclusions

The findings of this study allow us to conclude, for the first time, that mutual adjustment is a relevant concept in studying college graduates’ internships in small companies. Future studies, based on the operationalization of this concept, may foster the comprehension on how internships may result on job creation. This is even more critical, as internship programs have been used as an important youth employment policy in European Union countries in the context of the employment crises that have been occurring. These findings may be related to this internship programs’ characteristics of a regular on-the-job supervision, preventing interns from abandoning the internships. The relative low rate of internships turning into fulltime employment [21] suggests that the supervision of internships should be regular and conducted by independent supervisors, thus being relevant the program’ characteristics under analysis. This could justify other internship programs to include close on-the-job supervision if employment creation is the objective to accomplish. This study is also relevant for focusing on the capacity of small companies to create youth employment, since research on this type of company is rare and difficult to carry out and consequently little is known about them [67].

Finally, it may be concluded that organizational socialization practices and tertiary education sector directly influence mutual adjustment. This suggests that the institutional habitus of certain universities may influence how graduates adjust in an employment situation. As such, schools could also develop practices for monitoring internships and take into account how their students socialize in these contexts. Training sessions on organizational socialization practices could also be beneficial in facilitating the adjustment of graduates in professional contexts, particularly in small businesses. The most recent empirical studies emphasize that organizations invest heavily in orientation programs but tend to ignore the impact of supervisors who are “on the ground” with newcomers [13,68], as previous research also demonstrated [59]. Our study extends the literature and knowledge on mutual adjustment, also stressing the role of SE top managers’ perceptions which impacts youth employment. Finally, it is hoped that these findings will support researchers to further investigate how interns and SE top managers get involved in the mutual adjustment. As a last suggestion, it would be interesting to analyze to what extent participation in internship programs lead to processes of organizational change or to urge the internationalization of small companies.