Embracing public relations (PR) as survival panacea to private colleges’ corporate image & corporate identity erosion

Abstract

The research study explored the role of Public Relations (PR) elements such as seminars and speeches, digital media in creating a positive corporate image and corporate identity and explored how selected private colleges in Chitungwiza, Zimbabwe can leverage on publicity to boost students’ enrolment. Extant literature of PR was approached and the Hedging and Wedging theory of communication was used as the theoretical framework for the research. The researcher applied mixed-pluralist research approach, where the qualitative research design was nested into partially quantitative research design. Methodology applied was exploratory, deductively anchored on the critical analysis of the data and discussion of findings to arrive at an informed decision. A purposive sample of 300 students, 30 teachers, 10 college directors and 30 administrators and self-administered closed-ended questionnaires were used in different private colleges. Structural Equation Modelling (SEM) using AMOS version 6 software, was used to test the propositions. Major findings are that seminars and speeches are less frequently than other digital PR approaches done in private colleges and this can affect the students’ enrolment and survival of the former. Recommendations included the need to make an improvement on carrying out frequently on the seminars and speeches, so as to enhance public image of the former. Budgeting for the PR function in private colleges is crucial. Students need to be involved in the publicity activities of a private college.

Keywords:

• College marketing communications
• educational business management
• students’ enrolment
• strategy
• public relations

Subjects:

• M39
• M310

PUBLIC INTEREST STATEMENT

Publicity and Public Relations (PR) should not mixed. It is difficult to differentiate them due to the little difference. PR and Publicity are intertwined but they have different perceptual rating by the people. Publicity is viewed in a less good way than Public relations. If the publicity is not handled properly, it may result in negative effects to the institution, to which it is concerned to. Private colleges should be in a position to properly take enough scrutiny when handling the two coherent areas so as to minimize problems when dealing with corporate image and identity issues. The general beauty, the lighting, ablution services, the landscape, the ambience and the way the private colleges are structured plays an important part in attracting the enrolment of students. The research contributes to theory, practice and future research.

1. Introduction & background analysis

The private education sector in Chitungwiza (second largest city in terms of population statistics in Zimbabwe) has been facing difficulties since its inception as a result of resilience lack and unwavering competition fostered by the public sector educational institutions (those owned by the central government and local authorities). The main aim of the study was to explore the impact of Public Relations (PR) activities on students’ enrolment in selected private colleges found in Chitungwiza. Private learning institutions mushroomed in Zimbabwe over the past decade as the country’s public education system collapsed. The PR element within the selected private colleges in Chitungwiza, Zimbabwe, have been used with little significance and partiality. This has resulted in leading the former to succumb to the pressure of competition rivals and unethical practices. The major problem being the negative perceptions by the public, related to services offered in the private education sector. This could be lack of honesty and trust in service delivery, poor professionalism and adherence to standards, no positive working relations with the stakeholders, charging of exorbitant fees without commensurate value for money services, high staff attrition leading to instability of tenure and others. Further to this leakages of exams, external assistance and smuggling of foreign materials (in hems, pants or body parts) are some of the causes. Zimbabwe has continued its crackdown on unregistered colleges and private schools, closing down some 160 uncertified learning institutions. One of major factors negatively affecting public relations and image of these private colleges is the examinations leaking issue. The preliminary investigations from the management of a few private colleges in Zimbabwe, showed that they are using several marketing programs to retain students available in a move meant to improve their competitiveness (Mashanyare & Chinamasa, 2014), in terms of reviving dimuniting corporate images and corporate identities. Further to this, it is not only peculiar to Zimbabwe alone. In Egypt, it is alleged that there was live examinations’ leaking and mix up during the distribution process, as a result of mix up before they had been written (Clark, 2012; Mashanyare et al., 2014). Again, in South Africa a school principal was killed by militia on his way to collect examinations papers, leading to compromise of the security concerns (Otsialo, 2013). In reference to The Zambian Watch Dog (2013), Zambian Chemistry (5124/2) and Mathematics (Syllabus D) for 2012, were leaked through senior officers, who managed to buy the papers for their children at Kabulonga and Munali boys colleges. In addition to the above, 124 private colleges and independent training institutions were shut down while 31 others were de-registered when they failed to comply with the country’s education laws, which have affected the public relations and perceptions towards the former. With this comprehensive background, the research study answers the following research questions: What are the roles of seminars and speeches on students’ enrolment? To what extent does digital media impact on creating a positive corporate image and identity for private colleges? What are the ways in which private colleges can leverage to gaining competitive advantage? Before delving into the preceding, the following is brief literature review relating to the research study.

2. Literature review

Through the use of digital media many of the people with their homes are now more access to reach majority of their target population the use of digital media platforms. The television sets, radio transmission, either online or conventionally, they can have some access as they do it. The use of digital media platforms is a comfortable and profitable way to reach the targeted audience. This same applies as well to the private college in Zimbabwean education sector. They can use the digital media, to create a good positive image in leveraging on the mass media digital platforms. It is very costly to keep abreast with the ever dynamic digital media platforms. The entails the colleges as they operate, they need to focus their efforts and concentrate on the digital media as a way of improving the enrolment figures, easily and perfectly (Figure 1). Having digital platforms, it is not only limited to the traditionally used ways but also to use even texts, voice messages, and other contemporary digital media tools (Hein, 2004; Morley 2016). Karadeniz (2009) put forward that a variety of the instruments such as lobbying are a good blend or mixture for the essence of making publicity an important aspect in building a good name for an institution like a private a college. Kerin et al. (2006), in Karadeniz (2009), said Publicity and public relations should not mixed. It is difficult to differentiate them due to the little difference. Although this might be the case, diverse views have been made by authors like Karafakıoğlu (2005), who argues that PR and publicity are intertwined but they have different perceptual rating by the people. Publicity is viewed in a less good way than Public relations. Karafakıoğlu (2005), further deepens his arguments by simply saying if the publicity is not handled properly, it may result in negative effects to the institution, to which it is concerned to. This view means that private colleges should be in a position to properly take enough scrutiny in handling the two coherent areas so as to minimize problems. The general beauty, the lighting, ablution services, the landscape, the ambience and the way the private colleges are structured plays an important part in attracting the enrolment of students (Bell, 1999). Private colleges’ decoration exterior and interior design work a lot in marketing and branding efforts (Oplatka, 2002a, 2002b). Most of the schools in England tend to focus on the nature of the beauty of its entrances (Bell, 1999; Bell & Bell, 1976). Having the clubs and extracurricular activities in the area of decorations and design, it tends to improve the way the learners can take hold of their actions, self-control and respect as well as increase the room for total commitment to the school. Zhuldz et al. (2013), converge their views with Wong et al. (2009) in that, if an organisation is able to improve its link of sponsoring other firms or the public, the ultimate result would be an increase in the output, and shaping of reputation of a concerned party. Wong et al. (2009), suggest that engagement of a firm or institution into charitable deeds, activities of the societal benefits, can earn it a good positive image in a myriad of competitors. This entirely mean private colleges can leverage on the donations and corporate social responsibility activities to build good brands for themselves. The author believes that campaigns are a necessary publicity tool that can be used by private colleges to build trust and loyalty to their brand (Nyagadza et al., 2020a, 2020c, 2020d, 2020b). Many of the institution who have or once been into campaigns have gained a resultant benefit of these issues in their organisations. Kendall (1992), as cited in Niemann and Grobler (2001), believes a campaign is an organised activity, having more precision than a mere program, because it does have a clear picture of its intended purpose and control over a given objective. Private colleges, can be in a strategic position to leverage on the campaigns, as they sound to be more effective tools in gaining more publicity than any other valuable comparison tools (Kendall, 1992; Niemann & Grobler, 2001). All the PR activities lead to recovery of corporate identity, which is viewed as ethos, aims and values that create as sense of individuality (De Chernatony, 1999; Harris & De Chernatony, 2001; Harrison et al., 1998; Kapferer, 1997; Van Riel & Balmer, 1997) which helps to differentiate a private college’s brand from others. The concept of corporate image, reputation, branding and identity has been discussed highly by scholars in various disciplines (Balmer, 2001; Caruana, 1997; Eberl & Schwaiger, 2005; C.J. Fombrun, 1998; Harris & De Chernatony, 2001; Helms, 2007; Knox & Bickerton, 2003; Hatch & Schultz, 2003).

Figure 1. Private colleges’ public relations corporate image & corporate identity reputation reconstruction conceptual model

Source: Researchers’ conception (2021), derived from literature and theoretical review

Stamm and Grunig believe that attitudes can be moved from wedging to hedging to wedging, but not wedging to wedging (Skinner, 1994). The implication of this for public relations programmes is significant, as the intention is often to diametrically change people’s views about an individual or organisation (Skinner, 1994). A conceptual framework is an analytical tool with several variations and contexts. It was used to make conceptual distinctions and organize ideas. Strong conceptual frameworks capture something real and do this in a way that is easy to remember and apply (Makasi, 2015). The construction of the framework may take various forms highlighting the main variables of interest to the researcher and how they are interlinked through the use of circles combined with narrations to convey meaning and rationale of the framework (Hamilton & Selen, 2004; Makasi, 2015). In sync to the preceding research model, the simple regression model below was used to determine linear relationships between variables:

$Y={\beta }_0+{\beta }_1{X}_1+{\beta }_2{X}_2+{\beta }_3{X}_3+{\beta }_4{X}_4+u_t$

Where Y = Private Colleges’ Corporate Image & Corporate Identity Reconstruction,

$X_1$ = Seminars & Speeches,

$X_2$ = Social Media Networks,

$X_3$ = Radio & TV Broadcast,

$X_4$ = Publicity Activities,

$u_t$ = Unobserved error term,

*Y is the Dependent/Response Variable.

*From $X_1$ to $X_4$ represent the Independent/Explanatory Variables.

The mediator variable of this research was Hedging and Wedging, the dependent variable Private Colleges’ Corporate Image & Corporate Identity Construction and independent variables included Seminars & Speeches, Social Media Networks, Radio & TV Broadcast and Publicity Activities.

3. Research propositions

The following five (4) hypotheses were developed from theoretical and literature review;

$H_1$: There is a positive relationship between Corporate Image & Corporate Identity and Seminars & Speeches.

$H_2$: There is a positive relationship between Corporate Image & Corporate Identity and Publicity Activities.

$H_3$: There is a positive relationship between Publicity Activities and Social Media Networks.

$H_4$: There is a positive relationship between Corporate Image & Corporate Identity and Radio & TV Broadcast.

The next section discusses methodological issues applied in the current research study.

4. Methodological delineations

The researcher applied mixed-pluralist research approach, where the qualitative research design was nested into partially quantitative research design. Methodology applied was exploratory, deductively anchored on the critical analysis of the data and discussion of findings to arrive at an informed decision. Structural Equation Modelling (SEM) using AMOS version 6 software, was used to test the propositions. Researchers demonstrated their axiological, ontological and epistemological skills by being able to articulate their values as a basis for making judgements about what research they are conducting and how they go about doing it (Cohen & Manion, 1996). In conducting the research, it was made up of a sample of 10 private colleges’ directors, 30 educational administrators, 300 students from different colleges and 30 educational practitioners (teachers). The justification for this is due to the fact that rest of the private colleges in the area of study were not registered, so dealing with a few registered can yield tangible formal results. For sampling, Purposive technique was employed to reduce errors and biases. The question items were of Likert scale, which rated impact of PR through strongly agree, agree, not sure, disagree and strongly disagree (David, 2009; Vutete & Chikosha, 2014). A self-administered closed-ended questionnaire was used to collect data.

5. Data analysis

Both descriptive and inferential statistics were used in analysing quantitative data from the questionnaire. Structural Equation Modelling (SEM) was used to test the posited hypotheses. Descriptive statistical analysis was achieved through the functional application of charts, tables, graphs and diagrams, and this fed into inferential statistics. These included frequencies, mean, and standard deviation. Software packages used were Statistical Package for the Social Sciences (SPSS) and Analysis of Moments of Structure (AMOS) version 6. Exploratory Factor Analysis was applied to reduce the data collected into its underlying themes (Vingirayi et al., 2020) in assessing corporate storytelling as a corporate reputation cue. To assess adequacy of the measurement model, the researchers applied Confirmatory Factor Analysis (CFA). The researchers also utilised Principal Component Analysis (PCA) to consider the total variance in the data and establishing minimum number of factors that will account for the maximum variance. In addition, the Bartlett’s test of sphericity was applied to examine the hypothesis that the variables were uncorrelated. It was used to see whether there were some relationships between variables, which is necessary for factor analysis to be appropriate (Field, 2009).

6. Reliability & validity

In this investigation, reliability and validity was promoted through asking similar questions to the external teachers, students and experts outside the same particular private colleges. The Delphi was used to validate the items for measurement. External teachers, students and experts from private colleges, agreed to participate. On generalizability, the research’s qualitative findings can be used to infer to the underlying population since the research data deployed covered private colleges in Chitungwiza City, Zimbabwe. Reliability of each factor was tested using Cronbach’s alpha. Each value was required to be at least .5, as this is suggested to be a sufficient reliability score by (Makasi, 2015). Internal consistency was meant to measure the degree of interrelatedness of measurement items that were constructed to assess the uniformity. To assess validity, content, discriminant and predictive validities were tested. The researchers used content validity to look into the fitness and link of the research subjects to the theoretical underpinnings. Furthermore, researchers employed pre-testing and pilot approaches to enhance research instrument’s content validity. The concept of construct validity used was made to check on the connections between items that were assessed and the concept under study. To assess construct validity, average inter-item correlations were computed using CFA. To establish discriminant validity of the measurement model, the researchers employed Fornell and Lacker’s, measure of Average Variance (AVE). All the factor loadings that were above .5 were considered. Researchers’ reflexivity was employed so as to enhance trustworthiness of the inquiry, by disclosing all personal beliefs and values that underpinned the study. To safeguard the skipping of important information, the data was analysed following an iterative format, in addition to member validation. The process of debriefing was done and an analyst triangulation was utilised, where an audit trail to check for adequacy of the research process was also involved (Nyagadza, 2020a, 2020b, 2020c). To search, screen and select literature for review and research results analysis purposes, the databases Taylor & Francis, Scopus, Web of Science, Elsevier, Wildcat, Sabinet were consulted. Keywords such as PR, publicity, education, social media guided the search process. The query included peer-reviewed journals heralded between January 2010 and November 2020.

7. Sufficiency & saturation

The number of participants was not determined prior to the research. The criteria of sufficiency and saturation was used to determine the number of participants who responded in the study (Chiome, 2012; Rossman & Rallis, 2003). The point of saturation was reached when the researcher began to get the same information being reported by additional participants. Sufficiency was reached when enough people were interviewed to reflect the range of participants and issues for the study (Chiome, 2012; Rossman & Rallis, 2003).

8. Non-Response bias test

Armstrong and Overton’s techniquewas used to check for non-response bias test. The process involved the use of t-tests to compare the means of each of the items of the succeeding responses against the rest of the responses. There were no larger differences in the means. This suggests that non-response bias was not a threat to the research study.

9. Ethical considerations

Working on any information connected to surviving individuals made personal data processing essentially restricted and was subject to the provision of Data Protection 1998 (Nyagadza et al., 2020a). No third parties were allowed to gain access to the data collected (Marketing Research Society (MRS), 2020). Possibilities for gaining the room for security, anonymity and confidentiality were considered. Despite institutional safeguards, the burden for conducting ethical research lies with the researcher (Chiome, 2012; Merriam, 1998), participants were assured of the former. Each participant was given a statement of confidentiality with the provision for non-participation or withdrawal from participation at any time.

10. Respondents’ profiling

A grand average response rate of 91.89% was achieved in the collection of data. Majority of the respondents approached responded positively. Teachers had 93.6%, administrators and directors had a 90% and 80% response rate, respectively, whilst that of students had higher rate of 96.6%. Table 1 depicts respondents’ sex disaggregation as well profiling.

Table 1. Interview & questionnaires respondents’ profiling

Classification	Sex	Disaggregation	Percentage	Age Range	Total	Response
Students	Males	147	49%	13 ≤19	300	96.6%
	Females	153	51%	13 ≤19
Teachers	Males	11	37%	28 ≤57	30	93.6%
	Females	19	63%	25 ≤60
Administrators	Males	17	57%	45 ≤62	30	90%
	Females	13	43%	45 ≤63
Directors	Males	7	70%	45 ≤63	10	87.7%
	Females	3	30%	45 ≤63

Source: Field data (2020)

The teachers who responded 67% had a zero less than or equal to five-year period with the private colleges, 18% had plus six less than or equal to 10 years and 15% had 10 years or more. Student respondents 20% of them had a zero less than or equal to 1-year period with the private colleges, 50% had plus two less than or equal to 4 years and 30% had 5 years or more. The sample was only made of 300 students, 30 teachers, 30 administrators and 10 college directors from Chitungwiza, Mashonaland East province, Zimbabwe, who were willing to respond to the questionnaire. All were from secondary private colleges. Succeeding section accounts for results analysis and discussion.

11. Results analysis & discussion

Both descriptive and inferential statistics were used in analysing quantitative data from the questionnaire.

12. Sample adequacy

Both the Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy and Bartlett’s test of sphericity were conducted and the findings from the tests are displayed in Table 2. In this research, the two tests of sphericity were used. The Kaiser-Meyer-Olkin (KMO) Indexes of sampling adequacy for all factor analysis were explored using KMO. This is to ensure the sufficiency of covariance in the scale items to warrant factor analysis.

Table 2. KMO and Bartlett’s test

Kaiser-Meyer-Olkin Measure of Sampling Adequacy.		.985
Bartlett’s Test of Sphericity	Approx. Chi-Square	2351.175
	df	21
	Sig.	.000

Source: Primary data (2020)

According to the results in Table 2, the probability value for the KMO test is .985 which is above the threshold value of .5. This means that the relationship between the items is statistically significant and is appropriate for EFA to produce parsimonious set of factors. The Bartlett’s test of sphericity p < .05, implying that the correlation matrix is not an identity matrix. The results indicated that the matrix was not an identity matrix and this allowed the factor analysis to be conducted as relationships between variables existed. Sample size is another crucial aspect that should be considered prior to meaningful analysis because it relates to the stability of the parameter estimates. Based on the results of Table 3 the Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy was .935 for all constructs, above the commonly recommended value of .600. Bartlett’s test of sphericity (test of at least one significant correlation between 2 of the items studied) was also significant (χ2 (105) = 2351.175, p < .05) indicating that the correlation matrix was suitable for factor analysis. The communalities (proportion of item’s variance explained by the extracted factors) were all above 0.300, further confirming that each item shared some common variance with other items.

Table 3. Normality tests

Variables	Kolmogorov-Smirnov^a			Shapiro-Wilk
	Statistic	df	Sig.	Statistic	df	Sig.
Seminar & Speeches Social Media Networks Radio TV & Broadcast Publicity Activities	.132 .113 .144 .128	340 340 340 340	.000 .000 .000 .000	.919 .928 .927 .932	340 340 340 340	.000 .000 .000 .000

Source: Primary data (2021)

13. Normality tests (Kolmogorov–Smirnov (K-S) and Shapiro–Wilk (S-W))

The Kolmogorov–Smirnov (K-S) and Shapiro–Wilk (S-W) normality tests were conducted to examine if all variables are normally distributed prior to conducting the parametric tests (Field, Miles & Field, 2012; Shapiro & Wilk, 1965). The test was done under the null hypothesis which suggested that data from the sampled variables were normally distributed. These tests were conducted in order to inform the decision on whether to use parametric or non-parametric statistics. Hypothesis test for a test of normality:

H₀: The data is normally distributed

H₁: The data is not normally distributed

Data normality was assessed using Shapiro–Wilk’s test (as shown in Table 3). Results show that data were approximately normally distributed. Shapiro-Wilk’s tests were statistically insignificant for all latent constructs (p > .05) (Field, Miles, & Filed, 2012; Shapiro & Wilk, 1965).

13.1. Reliability & validity analysis

Pre-testing using Cronbach’s alpha test was done to test the reliability of the questionnaires. This helped to pre-emptively highlight any misleading questions or confusing questions that would distort the participant’s responses thereby contaminating the results. Table 1A (Appendix A) is a summary of the descriptive statistics and the Cronbach’s alpha values for the items, the standardised factor loadings for all items, Composite Reliability (CR) and Average Variance Extracted (AVE) that measure construct validity (Churchill, 1979). The quality of this research instrument used for the purpose of this study, the reliability measurement was carefully tested.The widely accepted social science cut-off point, alpha value should be 0.70 or higher for a set of items to be considered a scale, but some use 0.75 or 0.80, while others are as lenient as 0.60. In order to analyse the interrelationships among large numbers of variables, and to describe such variables in terms of their common underlying factors, Exploratory Factor Analysis (EFA) is used.All the factor loadings and the construct reliability for convergent validity should be equal to or greater than .7 for good convergent validity. In this case, all the factor loadings, Composite Reliability (CR) and Average Variance Extracted (AVE) that measure construct validity were above this threshold. This showed that the convergent validity conditions were all met (as depicted in Table 5). According to the results displayed in Table 3 there is Cronbach’s alpha internal consistency test for all the sub-dimensions and the scale which is composed of the item in total gave the following results: (Seminar & Speeches) $\alpha =.996$, (Social Media Networks$)\alpha =.997$, (Radio & TV Broadcast) $\alpha =.996$ and (Publicity Activities) $\alpha =0.995$. (Corporate Image & Corporate Identity$)\alpha =.998$. These results indicate that the survey instrument used showed high evidence of reliability. Another measure of internal reliability employed in this study is composite reliability. The composite reliability values for all constructs were between .990 and .998 which is generally acceptable, signifying the attainment of satisfactory levels of composite reliability.

Table 4. Regression coefficients

Model	R	R-Square	Adjusted R-Square	Std. Error of the Estimate
1	.994^a	.988	.988	.224

^a.Predictors: (Constant), Publicity Activities, Seminar & Speeches, Social Media Networks, Radio & TV Broadcast

Table 5. Regression coefficients

Model	Unstandardized Coefficients		Standardized Coefficients	t	Sig.	Collinearity Statistics
	B	Std. Error	Beta($\beta )$			TLR	VIF
(Constant) Seminar & Speeches Social Media Networks Radio & TV Broadcast Publicity Activities	−137 .114 .480 –041 .471	.044 .038 .087 .100 .070	.110 .467 -.040 .465	−3.140 2.977 5.498 -.412 6.901	.002 .003 .000 .061 .000	.54 .010 .008 .016	18.516 97.995 128.259 62.028

^a.Dependent Variable: Corporate Image and Corporate Identity (CICI)

13.2. Multiple regression analysis

According to the multiple regression results, R-square value of .988 was noticed indicating that 98.8% total variability in the dependent variable was being explained by the independent variables. Furthermore, the R-square (.988) was significant, implying this was a valid regression model that can be used for prediction purposes.

All the regression coefficients (Table 4) in the model were statistically significant at 5% since all the p < .05. A multiple regression modelling technique is adopted to examine how the dependent variable (Corporate Image and corporate Identity) is influenced by each of the explanatory variables (Seminar & Speeches, Social Media Networks, Radio & TV Broadcast, and Publicity Activities). The regression results are displayed in Table 4.

All the regression coefficients in the model displayed in Table 5 are statistically significant at 5% since all the p-values are less than .05 except for radio and TV broadcast with p-value of .061 though it is very close to .05. The absolute value of $\beta$ in Table indicates the order of importance of the independent variables. The variable with the highest $\beta$ value is relatively most important independent variable. On examining the contribution made by the independent variables in the model, it was found that social media networks made the biggest contribution with the value of ($\beta =$.67). It was followed by publicity activities and seminar and speeches with .465 and .110 respectively. Although radio and TV broadcast contributed less to the model, it was entered into the model due to the property of the regression analysis. Based on the regression analysis results, the multiple regression equation was obtained as it is shown below:

$\mathrm{y}=-.137+.114x_1+.480x_2+0.-.041x_3+.471x_4$

Where: $x_1$ denotes Seminar & Speeches (SS), $x_2$denotes Social Media Networks (SNM), $x_3$ represent Radio & TV Broadcast (RR), $x_4$ and denotes Publicity Activities (PA).

It can be noted that an increase in seminar and speeches values by one unit will result in Corporate Image & Corporate Identity increasing by .114. An increase in Social Media Networks and Publicity Activities by one unit will result in corporate image and corporate identity increasing by .480 and .471, respectively, whilst other independent variables remain constant. However, an increase by one unit in Radio & TV Broadcast will result in Corporate Image & Corporate Identity declining by .041.

Variance inflation factor (VIF) and Tolerance were also examined. Multicollinearity is detected by examining the tolerance for each independent variable. Tolerance (TLR) represents the proportion of variation unexplained in independent variables. Values less than .1 may be considered to indicate the presents of multicollinearity. From the coefficient table, all the variables have values less than .1 except for seminar and speeches and this indicate presents of multicollinearity. A VIF greater than 10 can be considered indicating the presence of more severe multicollinearity. All the variables have VIF greater than 10 indicating severe multicollinearity though seminar and speeches has a lower value as compared to the other explanatory variables. Radio & TV broadcast has a VIF of 128.259 which is higher than the other three variables in the regression model. This explains why Radio & TV broadcast contributes less in the model. Multicollinearity can be resolved by combining the highly correlated variables through principal component analysis, or omitting a variable from the analysis that associated with other variable(s) highly.

14. Analysis of Variance (ANOVA)

Analysis of variance (ANOVA) was carried out to see if the explanatory variables (Seminar & Speeches, Social Media Networks, Radio & TV broadcast and Publicity Activities) have the same impact on the Corporate Image & Corporate Identity. Denoting the coefficient of Seminar & Speeches by ${\beta }_1$, social media networks by ${\beta }_2$, Radio & TV Broadcast by ${\beta }_3$ and the publicity activities by ${\beta }_4,$ the ANOVA was conducted under the following hypothesis:

H₀: ${\beta }_1={\beta }_2={\beta }_3={\beta }_4=0$

Versus

H₁: ${\beta }_i\ne 0$ for at least one $i\left(i=1,2,3,4\right)$

The following Table 6 is a summary results of the ANOVA test from the research data.

Table 6. Summary the results of the ANOVAa test

Model		Sum of Squares	df	Mean Square	F	Sig.
1	Regression	678.247	4	169.562	3380.458	.000^b
	Residual	8.276	165	.050
	Total	686.524	169

a. Dependent Variable: Corporate Image & Corporate Identity

b. Predictors: (Constant), Publicity Activities, Seminar& Speeches, Social Media Networks, Radio & TV Broadcast

Source: Primary data (2021)

An examination of Table 6 makes it clear that the four independent variables in the standard model are significantly predictive of the dependent variable CICI according to the ANOVA statistics [F(4;165) = 3380.458, p < .05]. The next section introduces Structural Equation Modelling (SEM) in full detail, explaining how hypotheses were tested.

15. Structural Equation Modelling (SEM)

Structural Equation Modelling (SEM) was used to test the relationships amongst the variables in the model. Multiple regression analysis of factors among a single measured dependent variable and predictors group have been applied. This was meant to show the relationships between constructs and causal dependencies between exogenous and endogenous variables. Therefore, the SEM in this study was chosen to maintain parsimony in the proposed model, while benefiting from its strength in testing the research hypotheses. The relationships between the independent variables; moderator variable and one mediator was examined. Observable and latent variables in the model, were used to execute full Structural Equation Modelling (SEM).

16. Structural & measurement model

Structural Equation Modelling (SEM) integrated both path and factor analyses including multiple observed variables and indicators for every unobserved variable as in factor analysis and specific paths connecting the unobserved variables in the path analysis (Bagozzi & Yi, 1988). Confirmatory Factor Analysis (CFA) established a model to fit to the data. Analysis of oment structure (AMOS) was used to examine and quantify the relationships between independent and dependent variables. The model was then evaluated using Goodness-of-Fit (GoF) test and hypothesis testing. A graphical presentation of the model is displayed in Figure 2.

Figure 2. Initial model with standardized estimates

Source: Primary data (2021)

Codes’ Key for Figure 2: PA—Publicity Activities; SS—Seminars & Speeches; SNM—Social Media Networks; RR—Radio & TV Broadcast and CICI—Corporate Image & Corporate Identity.

From Figure 2, if Publicity Activities (PA) is increased by one, Corporate Image & Corporate Identity (CICI) will increase by .36. If Radio & TV Broadcast (RR) is increased by one, Corporate Image & Corporate Identity (CICI) will increase by .87 whereas Corporate Image & Corporate Identity (CICI) will increase by .17 when Social Media Networks (SMN) is creased by one. Moreso, if SS is increased by one, Corporate Image & Corporate Identity (CICI) will increase by .07. The next section explains model fit analysis.

17. The model fit analysis

Researchers used incremental fit indices as well as absolute fit indices. The Comparative Fit Index (CFI) and Root Mean Square Error of Approximation (RMSEA) provide sufficient special information to evaluate the model (Hair et al., 2010; Foroudi, 2018). The model fit analysis summary results for various fitted models are displayed in Table 7.

Table 7. The model fit analysis summary results

Model	χ^2	DF	PCMIN/DF	GFI	TLI	CFI	NFI	RMR	RMSEA
1	11,962.701	556	21.516	.414	.762	.777	.769	3.120	.246
2	3195.224	455	2.852	.862	.930	.946	.938	.016	.078

Source: Primary data (2021)

18. Goodness-of-Fit (GoF) statistics for final model

The CMIN/DF was not exceeding the threshold value of 3, implying that it was a good fit. The Goodness-of-Fit Index (GFI) value of .862 is slightly less than .9 indicating a better fit while both the Tusker Lewis Index (TLI) (.938) and Comparative Fit Index (CFI) (.946) values were all above threshold value of .9 indicating a very good fit and is considered to be the improved version of NFI index (Hair et al. 2006). GFI index measures the fitness of a model compared with another model (Hair et al. 2006). CFI is an incremental index that measures the fit of a model with the null baseline model (Hair et al. 2006; Foruodi, 2018). TLI alternatively known as the Non-Normed Fit Index (NNFI), compares the χ² value of the model to that of the independence model and takes degrees of freedom for the model into consideration. The Normed Fit Index (NFI) value met the minimum recommended value of .9, implying it was a good fit, although it does not control for degrees of freedom and underestimates fit in small samples (Hair et al. 2006). RMR value of about .05 or less indicates a close fit of the model and in this case the RMR value was .058 which was reasonable considering that real data was used. There was a reasonable error of approximation as indicated by the Root Mean Square Error of Approximation (RMSEA) value (.078) which was below (Table 8) the recommended value of .08(Makasi, 2015).

Table 8. Goodness-of-fit statistics for the final modified model

Index	Recommended value	Sources	Final Model
X^2	-	Hair et al. (2010)	3195.224
DF	-	Hair et al. (2010)	455
PCMIN/DF	≤ 3	Hair et al. (2010)	2.852
GFI	≥ .90	Hair et al., (2010)	.862
TLI	≥ .90	Hooper et al. (2008)	.930
CFI	≥ .90	Hair et al. (2010)	.946
NFI	≥ .90	Hair et al. (2010)	.938
RMR	< .05	Hair et al. (2010)	.016
RMSEA	≤ .08	Hair et al. (2010)	.078

Source: Primary data (2021)

The measurement model was nomologically valid and acceptable. Standardized regression estimates allowed to examine the direct link between the research constructs. Table 9 presents the results.

Table 9. Standardized regression estimates of the final modified model

Construct	Estimate	S.E.	C.R.	P
CICI <—SS	.257	.034	7.523	***
CICI <—PA	.430	.119	3.625	***
CICI <—RR	.173	.133	1.307	.191
CICI <—SMN	.165	.124	1.333	.182

Source: Primary data (2021)

19. Results discussion

The Critical Ratio (CR) and the p-value are commonly used when testing statistical significance of SEM components. The first hypothesis sought to establish the nature of the Corporate Image & Corporate Identity and Seminars Speeches. The results in Table 9 suggest the acceptance of the alternative hypothesis which states that there is a statistically significant positive relationship Corporate Image & Corporate Identity and Seminars & Speeches.

The second hypothesis sought to establish the nature of the relationship between Corporate Image & Corporate Identity and Publicity Activities. In with the results in Table 5 there is a positive and statistically significant relationship between Corporate Image & Corporate Identity and Publicity Activities. The result suggests the acceptance of the alternative hypothesis; hence, we can conclude that there is a statistically significant positive relationship between Corporate Image & Corporate Identity and Publicity Activities. Karadeniz (2009) seems to be in a divergence by suggesting that publicity and related cases can be strategic tools used by organisations to improve the PR goals and objectives. However, the reason why there was neutrality with a considerably high percent could be due to the reason that some respondents were indifferent and some knew their private colleges were not normally involved in sponsorship (Kabonga et al., 2021). Once good publicity has been earned, this leads to a positive brand preference and perceptions within the stakeholders of the private colleges.

However, although this might the research result, on the contrary sponsorship is also regarded as tarnished activity because sponsors have exclusive right of propaganda in the sponsored affair or activity, which increases the possibility of using commonweal, charged Rodgers, Cameron and Brill (2005), as referenced by T Wang et al. (2008). Implementing the publicity activities with the inclusion of students, requires new roles of teachers to teaching and learning, including peer mentoring and experts demonstration of technical aspects (Cochrane & Narayan, 2012; Gan et al., 2015; Hoyos, 2014; Van Den Beemt et al., 2019). Negative connotations and associations have been attached to publicity activities (Fall & Hughes, 2009).These include spin doctoring as an alternative model of PR. Others negative connotations to the PR activities include publicity, lobbying, damage control and propaganda (Budd, 1997; Coombs & Holladay, 2007; Dilenschneider, 1999;Fall, 2004; Fall & Hughes, 2009; C. J. Fombrun, 1996; Ihator, 1999; Pinsdorf, 1999). The increase in private colleges’ brand awareness, brand equity, brand reputation and thereby increase the probability that their services will be evoked by resource holders, affecting perceived brand value, and creating an image that influences the private colleges’ services are perceived (Cobb-Walgren et al., 1995; Edell, 1993; Fombrun & Van Riel, 2004; Herr & Fazio, 1993).

The third hypothesis sought to establish the nature of the relationship between Corporate Image & Corporate Identity and Radio & TV Broadcast. A positive but statistically insignificant relationship between Corporate Image & Corporate Identity and Radio & TV Broadcast exist. We therefore accept the null hypothesis which states there is no statistically significant positive relationship between Corporate Image & Corporate Identity and Radio and TV Broadcast. In reference to the management in-depth interviews that were done, the frequently emphasised fact by majority of the respondents, was to embrace the idea of using broadcast media (Radio and Television) as PR supporting tool. This could ultimately enhance good image for the private colleges. However, the cheapening of products offered by mass media such as Radio and Television to defend corporate image and corporate identity of a private college is determined by the production industrialisation and the generalisation of the consumption (Coman, 2007; Florescu, 2014).

However, research findings by Wright and Hinson (2009b), in USA showed that the emergence of blogs has changed how people interact (53% in 2007 and 69% by 2008) (numbers continues to rise even in future trajectorially forecasts). With this, it limited the room and chance for opting for these variables. With this, there are offer some tips and suggestions to organisations and individuals on how to use the new media technologies in harnessing effective public relations. The “Other” online PR tools used for students enrolment was not specified by the respondents (Nyagadza, 2020a, 2020c, 2020d). This was as result that could be due to lack of knowledge or ignorance within the former. They ended responding in the easiest possible way to them. Research suggests that new online media platforms are having great impact on public relations (Tancer, 2008; Wright & Hinson, 2006a, 2006b, 2007a, 2007b, 2008a, 2008c, 2008d, 2008e, 2009 &, 2009b). To anchor this research result, there is a notion that sponsorship enhances the corporate image and corporate identity as cited in Tufail et al. (2014).

The fourth hypothesis sought to establish the nature of the relationship between Corporate Image & Corporate Identity and Social Media Networks. This shows that a positive but the relationship is not significant relationship. These results suggest the acceptance of the null hypothesis which states there is no statistically significant relationship between Corporate Image & Corporate Identity and Social Media Networks. One benefit of using social media to improve the image of the private colleges is that a larger audience can be reached, although this did not result in more student engagement (Bélanger et al., 2014; Cox & McLeod, 2014a, 2014b). This conclusively mean, respondents did not fully support the notion that, social media networks enhance positive image.

However, in support of this result, several scholars have researched the adoption of internet of things technologies and/or social media platforms for public relations activities (Dozier et al., 2016; Scott, 2010; Wang, 2015) and others have described this new development with terms such as digital, online or e-public relations (Amodu et al., 2019; Gifford, 2010; Petrovici, 2014; Philips & Young, 2009; Vercic et al., 2015). Although that might be the case, private college directors and administrators through the in-depth interviews; they argued that social media has some negative effects to some extent. Their views on this issue was saying, private colleges need to be cautious when dealing with issue of social media network, otherwise they can destroy a good image for a certain organisation or college. In support of this views by other scholars cite that social media as contributing to common online problems (Kahne & Bowyer, 2019) such as spread of misinformation (Kahne & Bowyer, 2019), echo chambers (Kahne & Bowyer, 2019; Pariser, 2011; Sunstein, 2007), and incivility (Coe et al., 2014; Kahne & Bowyer, 2019).

The results discovered that Facebook has more influence (topping at 35% in comparison to others) amongst the students, at Pamplin College of Business Marketing, Virginia, USA. This can also tally with the global statistics which tend to show that Facebook has a higher percentage in terms of the users. Twitter follows the hierarchy of respondents’ choice standing at 34%, Instagram and Google+ at 5%. The reason why the teachers have responded by opting less for the last two, it is a result that the social media platforms are still growing in terms of their usage and acceptability, as evidenced by the responses given in this ca. To support this, there is a suggestion that social media platforms have expanded opportunities for learning and interactivity. Furthermore, Q. Wang et al. (2011), suggest the younger generations of digital natives use technology, they not only exchange ideas, feelings and information (during school hours), but even visuals and money at a remarkable rate (Bahk, Sheil, Rohm & Lin, 2010).

20. Conclusion & practical implications

The researcher found that if private colleges in Chitungwiza City, Zimbabwe are to embrace the concept of PR with highest prioritisation, it may enhance a positive corporate image and corporate identity, which in turn increase the enrolment of the entire students. Embracing public relations leads to building of relationships, private colleges’ members (internal and external stakeholders), leverages current communications systems, and using different channels together have created an unparalleled, unified voice within the colleges. The recommendations have been deducted from the research study; Seminars and speeches are less frequently done in private colleges and this can affect the students’ enrolment. There is need to take improve on carrying out frequently on the seminars and speeches, so as to enhance public image of the former. Budgeting for the PR function in private colleges is a necessity. This will improve the visibility, creation of positive corporate image and corporate identity of the private colleges, which ultimately increase their enrolment numbers. Retention of the already enrolled students can easily be achieved through this. Sponsorship and donations are more of charitable deeds, then the more obvious their nature of educational business. Donations are more likely to make the public ascribe it to altruism motivation and motivation attribution. The authors suggests that students may need to be involved in the publicity activities of a private college. The Model UN or Mock trial can help the students’ engagement to social media publicity platforms and activities. The inclusivity improves enrolment and creates positive public relations and free publicity in shaping the private colleges’ corporate image and corporate identity. This is so because youthful private college students have better leverage power of online and digital publicity engagement (Mihailidis, 2018).

However, although the students are more involved in online or offline publicity activities than the older, there is also significant evidence that many of the students are largely disengaged from civic and political life and most of them rarely create and circulate political content, whether against their private college or nationally. Further to this, there is a major debate on media literacy educators regarding the degree to which media literacy education should focus on content from popular culture. The author charges that Radio and Television broadcast media are effective PR tools, although they are costly to run. Use of mass media such as the state run radio and television for private colleges’ public relations efforts enables a popular base for promotion of a positive corporate image and identity. Private colleges’ brand evaluations by stakeholders significantly is primarily related to publicity volume. This would make the stakeholders’ perceptions positive by viewing the private colleges’ corporate image and corporate identity more meaningful, more predictable and more trustworthy. The mass media can help build social consensus, enlightens (and avoid superficiality) only if there is an enabling environment. Educational administrators and/or managers for private colleges need to use it wisely and effectively. They can use the social media platforms to improve enrolment, public relations and publicity, especially the highly subscribed like Facebook and Twitter. This is due to the reason that, in this research students have advocated the use of social media in attracting them for enrolment. However, there is need to harness and confront, potentially contribution of social media to common online problems such as spread of misinformation. There is need for the private colleges’ educators and administrators to conceptualize and examine ways that educational interventions might influence students’ ability to gauge the credibility of social media content, to develop critical PR media literacies and surf the loopholes and leverage the opportunities of PR social media dialogue).

21. Limitations

Main research limitations are of the small sample size and the fact that the analysis has been conducted only at an aggregate level regardless of size or geographical area of the sample. Only limited to a few PR theories and models relevant to the study focus yet there are so many of them available for use.

Acknowledgements

We wish to thank the private colleges’ directors, educational management, practitioners and students of the institutions where the research was carried out, for their assistance.

Disclosure statement

No financial interest or benefit has arisen from the direct applications of this research study by the researchers.

Additional information

Funding

This work was self-sponsored by the authors.

Notes on contributors

Brighton Nyagadza

Brighton Nyagadza is a Doctor of Philosophy (Marketing Management) finalist (Midlands State University, Zimbabwe) and Digital Marketing Lecturer Marondera University of Agricultural Sciences and Technology (MUAST), Zimbabwe. He holds an Executive Certificate in Export Marketing & International Trade (University of Zimbabwe), Post Graduate Diploma In Education (Zimbabwe Open University), Professional Diploma in Marketing (Driving Innovation) (The Chartered Institute of Marketing, United Kingdom), Bachelor of Commerce (Marketing), (Zimbabwe Open University) and Master of Commerce (Marketing Strategy) (Midlands State University, Zimbabwe). His key interests revolves on corporate storytelling, marketing metrics and digital marketing.

Gideon Mazuruse is a full time Mathematics and Statistics lecturer under the Teaching & Learning Institute of the Marondera University of Agricultural Sciences and Technology (MUAST), Zimbabwe. He holds an MSc in Statistics and Operations Research (NUST), BSc Honours in Statistics and Operations Research (GZU), BSc in Mathematics and Computer Science (GZU), Post Graduate Diploma in Education (GZU).

Appendix A

Table 1A. Accuracy analysis statistics

Research Construct		Descriptive Statistics		Cronbach’s Test		CR	AVE	AVES/ Root	Factor Loadings
		Mean	SD	Item total	αValue
Speeches & Seminars	SS1	28.02	13.22	.988	.996	.998	.984	.988	.981
	SS2			.984					.976
	SS3			.989					.985
	SS4			.984					.975
	SS5			.985					.974
	SS6			.975					.956
	SS7			.983					.977
Social Media Networks	SMN1	30.92	181.39	.982	.997	.99	.983	.992	.970
	SMN2			.990					.984
	SMN3			.985					.977
	SMN4			.990					.987
	SMN5			.991					.985
	SMN6			.992					.989
	SMN7			.988					.980
Radio & TV Broadcast	RR1	30.89	183.28	.984	.996	.997	.984	.994	.972
	RR2			.993					.988
	RR3			.996					.993
	RR4			.988					.984
	RR5			.995					.992
	RR6			.994					.990
	RR7			.994					.991
Publicity Activities	PA1	30.29	173.7	.992	.995	.997	.967	.983	.990
	PA2			.989					.980
	PA3			.990					.988
	PA4			.993					.987
	PA5			.988					.981
	PA6			.988					.979
	PA7			.990					.982
Corporate Image & Corporate Identity	CICI1	30.02	189.78	.994	.998	.996	.984	.991	.989
	CICI2			.992					.986
	CICI3			.989					.985
	CICI4			.992					.989
	CICI5			.990					.981
	CICI6			.989					.985
	CICI7			.989					.984

Source: Primary data (2021).