Journalism Degree Motivations

Abstract

Scientific knowledge should reflect valid, consistent measurement. It is argued research on scale development needs to be more systematic and prevalent. The intent of this article is to address scale development by creating and validating a construct that measures the underlying reasons why undergraduate students seek a degree in journalism, the Journalism Degree Motivations (JDM) scale. Through a multimethod approach and seven-step process, a set of motivations that reflect existing theory and measures was developed. The JDM scale is composed of eight factors: social responsibility, reporting, social prestige, sports media, photography, writing, varied career, and numbers and science anxiety.

Keywords

journalism education motivations scale development factor analysis scale reporting

The goal of this study is to create a scale. We articulate some basic principles on scale development and reporting practices through the development a journalism degree motivations (JDM) measure. No validated scale exists measuring students’ motivations for a journalism degree, and researchers vary in their items used to measure it across studies. The intention is to use best practice suggestions from other fields to help guide researchers in the scale development and reporting process. Scales are “collections of items combined into a composite score and intended to reveal levels of theoretical variables not readily observable by direct means.”¹ Based on findings from several content analyses of journals including the field of communication,² scale development as a topic should be addressed across journals. The validity of findings from survey and experimental research is questionable if findings are based upon suspect measures. McCroskey and Young said they believed there was “flagrant abuse of factor analysis in published communication research.”³ And Kline argued, “It seems that many, if not most, factor analytic studies have at least one serious flaw.”⁴

Factor analysis is the most common approach used to investigate the dimensionality of constructs. Factor analysis is a technique that allows researchers to identify items that correlate or “hang together” to form a factor, while not producing high correlations on other variables.⁵ In factor analysis, there are many decision points, and those myriads of choices create opportunities for error. Content analytic research of scale practices in communication journals shows “users generally demonstrate insufficient understanding of factoring techniques and the implications of the various choices.”⁶ We attempt to dissect the steps by using data to illustrate the scale development and reporting process through the creation of a JDM scale. “Journalism degree motivations” is a fruitful operational measure for scale exploration because of the variations in items that have been used to measure it. And research in the area of vocational psychology is especially timely because the news industry is facing competition from alternative outlets.⁷ It may be helpful to have a scale that administrators can use to provide insight into how to best advise and serve students.⁸

No validated scale exists measuring undergraduate students’ motivations for a journalism degree. Motivations and vocational choices are constructs used to assess personality.⁹ To develop more operational clarity related to the motivations behind degree choice, a large item pool was reduced to propose a measure with scientific utility. A handful of qualitative studies have analyzed motivations,¹⁰ while quantitative work reflects variations across items used to measure motivations on questionnaires.¹¹ Several of these scholars have provided a foundation that leads to the next logical step: measurement development and construct validation.

Suggestions from methodologists primarily from the field of psychology were used to identify best practices to help develop the JDM scale.¹² In this article, the creation of this scale involved a seven-stage process: (1) a review of existing literature and survey items was conducted; (2) four focus groups were carried out with undergraduate students at two universities to determine whether additional items should be included; (3) feedback was sought to examine the validity of items; (4) a draft of the survey was pilot tested to determine how the data fell and to test the questionnaire; (5) a final survey was administered to undergraduate students at three large U.S. universities; (6) an exploratory factor analysis (EFA) was conducted to identify the structure of the scale, including the number of factors, and following analysis, the initial sixty-two-item scale was reduced to twenty-seven items; and lastly, (7) a confirmatory factor analysis (CFA) was used to test whether the scale was consistent with the proposed understanding of JDM.

Literature Review

It is common for scientists to not report whether or how they validated their latent measures. Operationalization, an important component of the theory-building process, is the process of connecting concepts to observations by developing procedures to measure variation in variables.¹³ One major purpose of theorizing is to encourage hypothesis testing by developing concepts with scientific utility.¹⁴ However, the extent to which any field relies on questionable measures is detrimental to the advancement of scientific understanding and prediction.¹⁵ Thus, other scholars should not assume the validity of published scales. Content analyses of communication journals prove that scholars have used improper procedures when creating scales.¹⁶ To address this issue, we hope to contribute to scholarship by using data to illustrate the instrument design process through the creation of a JDM scale.

JDM

Universities are often the first step in socializing people into the profession, while organizations are the second. One of the most important concepts in the field of education is motivation because the personal orientations behind their major choice can help researchers understand how motivations relate to variables such as performance, perceptions, and persistence.¹⁷ Journalism remains a popular academic major even as departments report they are remaking themselves in the face of structural, technological, and economic changes.¹⁸

However, the reliability and the construct validity of previous psychometric measures are uncertain. It is important to ensure observed variables measure what they intend to measure. It is challenging to build theory when researchers use different measures every time they examine the same construct: JDM. This practice means it is difficult to generalize and compare results across studies. Based on the literature review, most quantitative studies did not share the theoretical logic behind their inclusion of items used to measure the construct, and most scales present single-item measures to assess each degree motivation.¹⁹ Single scale items are poor measures of a dimension because they “are not likely to capture the complexity of most social variables consistently or comprehensively. Item responses are affected by abstraction, bias, cultural influences, response sets, trigger words, item order, and many other extraneous variables.”²⁰ Based on this potential bias, at least three items should be included in each subscale. Ideally, it is argued that each dimension should include four or five items.²¹ Another problematic aspect of most research is that factor analysis and reliability statistics were not provided to indicate the structural validity of instruments used to measure JDM.²² Due to this empirical gap, the goal of this study was to refine the JDM measure.

Journalism Degree Motivational Items

One initial step in the instrument creation process is to analyze how other researchers have both defined and measured the concept. The labels appearing in the literature are motivations, expectations, aspirations, reasons, and perceptions of a journalism career. Steven Chaffee stated, “Disciplined use of words encourages other scholars to employ the same terminology so that a growing body of research can be cumulative.”²³ The concept most represented in literature review concerning this line of research is motivations. It is assumed that people’s occupational choices are determined by particular motives. Splichal and Sparks defined motivations as “the specific factors or individual character (nature) determining the direction (goal) and intensity of his or her activity.”²⁴ This study is concerned with what arouses people’s interest in journalism prior to having been fully socialized into their occupation. As people appear to see a degree as an important asset, more research is needed to identify consistent reasons why people choose to major in journalism. The selection of a college degree program often reflects one’s occupational choice. People who get journalism degrees may or may not pursue journalism careers because journalism degrees provide skills that apply to a wide range of professions.

To determine how motivations were measured, the literature was coordinated to assess the reasons why people enter into a career of journalism. Based on the review of literature, a group of fifty-five items was created. An overarching theme appearing is that people choose journalism as a major primarily for intrinsic reasons. Intrinsic motivation means that people chose a particular path because they receive pleasure and satisfaction from their choice rather than selecting a major because of external forces such as pressure from parents.²⁵ Specifically, it appears that they chose journalism as a degree to master specific skills such as photography or writing.

Splichal and Sparks conducted the first international comparative research in a twenty-two-country study of 1,822 first-year students and their motivations to study journalism in 1986.²⁶ No significant differences were found among the students based on country. Authors developed and categorized items into four different areas: self-esteem, desire for personal awards, altruism, and other motivators.²⁷ The primary motivators that students answered yes to on the questionnaire was liking journalism as a profession, meeting interesting people, and a talent for writing. In the 1940s, dichotomous variables were recognized as biased, problematic measures because they did not effectively measure latent variables.²⁸ One way to address dichotomous variables is by computing similar variables creating item sums.²⁹

An open question was presented to first-year British and Spanish students asking them the primary reason for wanting to become a journalist.³⁰ Researchers divided up the responses concerning motivations into eight broad categories: (1) nonroutine, nonconventional, sociable, variety-filled, exciting, challenging, sociable nature of journalism; (2) most desirable/suitable occupation for me, interesting, satisfying job for me, interest in current affairs and news; (3) a creative occupation, love of writing; (4) scope for self-education; (5) good prospects and good income; (6) public service—generally expressed, (7) public service—reform/change society, campaign, investigate; and (8) other reasons. Response categories ranged from maybe to sure.

In the Hovden et al. study, first-year Nordic journalism students were asked to grade twenty motivations on a scale.³¹ The complete number of and description of points were not communicated in the paper. The authors’ table presented responses with rankings of very important and somewhat important.³² They reported the items reflected three broad categories: practical motives, idealistic motives, and personal motives. The results showed the primary motivators for entering the journalism field included a varied and lively work and the opportunity to work with interesting people, while becoming a celebrity ranked low. Similar to these results, Becker et al. surveyed U.S. journalism and mass communication students asking them to rank the reasons for entering journalism. Exciting career, the opportunity to write, and an interest in public affairs were rated somewhat high.³³

In a longitudinal panel study of three cohorts of Norwegian journalism students, Bjørnsen, Hovden, and Ottosen found variation and liveliness; to be creative, to take part in the discussion of social issues, and to work against injustice were reasons for entering the profession.³⁴ However, the master list of motivations was not published in the study. In a similar vein, Mellado et al. identified role conceptions of journalism students, which included factors reflecting a desire to promote democracy, support and watch government efforts, and connect information with the masses.³⁵

In Bowers’ study of University of North Carolina students, he asked journalism majors to rate journalism in comparison with other careers on several attributes, which included interesting work, useful to society, job opportunities, prestige, family life, and financial reward. Family life and financial reward were rated lower among students.³⁶

Burgoon et al. identified characteristics of journalism that would be considered attractive to minority students. The most influential factors were related to enjoyment derived from job, prestige, and activities such as writing and speaking.³⁷ Hanusch also found the desire to write was the most proportionately attractive attribute of journalism for Australian journalism students.³⁸ A small proportion of the respondents studied journalism because of their desire to change society, meet interesting people, and travel.

Weaver et al. asked working U.S. journalists, not students, why they chose a career in journalism. The qualitative responses were placed into four categories: writing, searching for answers, seeking truth, and serving the public.³⁹

Method

The development of this scale was carried out because evidence suggests that there are distinct reasons students choose to major in journalism. It is important to note that not all research is about testing hypotheses. Scale development research does not present hypotheses and research questions.⁴⁰ Instead, the intention of this study is to build theory at the operational level by identifying the dimensions of a latent construct, in this case, motivations for choosing to major in journalism.

Standard scale development procedures involve both qualitative and quantitative survey work to identify content for a scale and to validate its structure. The development of the initial set of items consisted of a review of a comprehensive list of motivations, expert advice, and focus groups with students majoring in journalism. Given a continuum between a representative national study that would be expected to receive a comparatively low response rate and a study of a single campus that would yield a high but unrepresentative sample, the decision was made to compromise, selecting five campuses for the study. The diversity of these five campuses, combined with a relatively high response rate, yielded a dataset that represented the best compromise between response rate and representation.

Focus groups were conducted with four groups of six to ten undergraduate student participants per group from two journalism programs in the southwestern and southeastern United States. Focus groups are useful for understanding how people feel or think about why they are seeking a journalism career.⁴¹ The goal was to identify potential items missing from the list of motivators not found in the literature and to check for internal validity. Following qualitative analysis of the transcripts of focus group video recordings, several items were added to the scale.⁴²

Next, a committee of researchers (i.e., international and national researchers who studied this topic, n = 6, and doctoral students, n = 3) scrutinized the scale stemming from literature, theory, and focus group data to assess content validity. That group of people was asked to assess internal validity by evaluating the items for redundancy and gaps. Based on feedback about the conciseness, grammar, and face validity, some items were reworded.

After expert advice was sought and focus groups were conducted, a pilot study was conducted with students (n = 104) from a different southeastern university in the fall semester of 2012 to establish face validity.⁴³ We added⁴⁴ and dropped items⁴⁵ after conducting an EFA in SPSS (version 21.0) to reveal any latent variables that may cause manifest variables to covary. The total number of items on the final version was sixty-two. When conducting an EFA, the larger the item pool, the better. It is important to ensure that most content areas are represented in an initial item pool.⁴⁶ EFA is useful for assessing theories and developing scales.⁴⁷ EFA is a complex technique because of the inductive reasoning that is involved.⁴⁸

To establish internal validity, the survey was administered to journalism undergraduate students in spring 2013 at three large U.S. universities in the Midwest, northeast, and southeast. The sample reflects a convenience sample because the survey required the permission and cooperation of administration officials. Students were asked to voluntarily participate through email invitations, flyers, academic advisers, and course instructors. Incentives included a chance to win one of forty-two $50 gift cards, and in a few cases, faculty colleagues offered extra credit for participation.⁴⁹

The survey was administered using Qualtrics, Inc. Web surveys were chosen because this national sample was most likely accessible via the Internet, which made this approach most suitable.⁵⁰ The survey was administered over a six-week period in Spring 2013, which included a week of spring break vacation for students. The questionnaire took approximately twelve minutes to complete. The majority of the items were based upon a seven-point Likert-type scale with the exception of a small number of demographic questions toward the end of the questionnaire.

Once partial responses were deleted, the journalism dataset consisted of 798 U.S. undergraduate university students (63.0% female, 26.3% male, 10.7% nonresponse) majoring in journalism with an average age of 21.0. Students were distributed across four tracks within journalism (54.5% broadcast, 31.1% print, 9.6% visual communication and photojournalism, and 3.8% online/digital/convergent). The respondents were 66.9% White, 10.9% Black or African-American, 5.1% Hispanic or Latino, 3.1% Asian, 0.4% Pacific Islander, 0.3% Native American or Alaskan Native, and 2.5% other. The survey had a 33.2% (N = 2,401) response rate for the three journalism programs. The dataset was separated in half to create two random datasets for the EFA and CFA due to the sufficient sample size. The second dataset was retained to confirm the factor structure identified in the first set. Several sample size recommendations exist, but a sample size of at least three hundred participants is suggested for factor analysis.⁵¹

To determine whether exploratory analysis was appropriate, the Bartlett’s test of sphericity was used to estimate the probability that there was an identity matrix with 1s in the diagonal and 0s in the off-diagonals. The identity matrix would be rejected if p ≤ .001. The Kaiser-Meyer-Olkin (KMO) is a measure of sampling adequacy. It compares the magnitude of the calculated coefficients with the magnitude of partial correlation coefficients.⁵² The KMO measures ranges from 0 to 1. It is suggested that values of .60 or higher are necessary to proceed with the factor analysis.⁵³

After confirming that the Pearson correlation matrix was factorable, the matrix was submitted for EFA. The responses were examined using principal axis factoring rather than a principal component analysis (PCA) for the EFA. Two of the most common critiques of scale procedures include the use of PCA and orthogonal rotation by researchers.⁵⁴ Despite recommendations to not use PCA when creating scales, the majority of researchers from multiple fields use it.⁵⁵ PCA does not discriminate between shared and unique variance, which leads to the overestimation of factor loadings leading researchers to retain too many factors. Principal axis extraction was chosen because it is less sensitive to weak loadings and more robust to nonnormal data compared with maximum likelihood.⁵⁶

Substantive factors were selected using a combination of the following criteria: parallel analysis (PA), Catell’s visual scree plot,⁵⁷ and theoretical convergence. PA was developed by Horn⁵⁸ and compares obtained eigenvalues against a randomly ordered dataset. Components were kept when their eigenvalues were larger than the values generated by the random data. Eigenvalues represent the amount of variance in all of the items that can be explained by a factor. The PA method has been recommended for determining an accurate number of factors to accept.⁵⁹ The visual scree plot was inspected and components were selected where the factors curve above the straight line drawn by the researcher.⁶⁰ The scree test has been shown to be effective when determining the number of factors to obtain.⁶¹ Promax with a k value of 4 was the oblique rotation method that was applied in this study.⁶² If a researcher is primarily interested in producing results that best fit the data, the factors should be rotated obliquely. Oblique rotation assumes that the factors are correlated, whereas orthogonal assumes that factors are uncorrelated to each other. Uncorrelated factors are not common in the social science fields. And orthogonal rotation also tends to overestimate loadings leading researchers to inappropriately retain or reject items, which in turn may cause problems when conducting CFA.⁶³

Item deletion is an expected part of the process. Simple structure is an approach for approximating what items to retain. Simple factor structure, as proposed by Thurstone,⁶⁴ was determined based on the following preestablished general criteria: item sets with item loadings at or above the .32 level,⁶⁵ no cross-loadings, no factors with fewer than three items,⁶⁶ and theoretical convergence⁶⁷ were considered when determining best fit to the data.

Following identification and interpretation of the factors, a subset of items was selected to represent the factors identified after confirming the internal consistency of the reduced-item scale and conducting a CFA using SPSS. To determine whether our hypothesized model fit, CFA using Amos 21.0 was conducted to determine how well our predicted model fit the theoretically driven item set.⁶⁸ No model perfectly fits the data. As a result, rule of thumb exists to assess whether the model is close enough. Fit can be assessed using several indices. At a minimum, the following information should be reported for the CFA: the chi-square test statistic with corresponding degrees of freedom and level of significance (p > .05), RMSEA (root mean-square error of approximation < .08) with its corresponding 90% confidence interval, and CFI (comparative fit index < .90) to determine the overall model fit.⁶⁹ Many researchers have criticized the inclusion of the chi-square test because of its sensitivity to large sample sizes. Thus, it is not considered a reliable indicator of fit even though it is traditionally reported.⁷⁰

Results

The data matrix was scanned for missing data, outliers, and linearity. The missing cases were assessed for randomness. It was determined that the missing cases did not display a clear pattern. Cases that displayed clear abandonment were deleted. Means ranged on the seven-point scale (strongly disagree to strongly agree) from 3.42 to 6.16 with standard deviations ranging from 0.906 to 2.270. Skewness ranged from −1.885 to 0.240, and kurtosis ranged from −1.541 to 4.235.

Bartlett’s test (χ² = 10,807.815, df = 1,891, p < .001) and the KMO statistic of .869 suggested that the correlation matrix was appropriate for factor analysis. The correlation matrix is available upon request. PA suggested eight factors and the scree plot suggested nine factors. As a result, seven, eight, nine, and ten factor solutions were sequentially examined. The seven, nine, and ten factor solutions were rejected because there were cross-loadings (i.e., loadings, or pattern coefficients, were salient on more than one factor), theoretical convergence, and an inadequate number of loadings. The eight-factor solution was retained because it best met simple structure.⁷¹ As shown in Table 1, reliabilities (coefficient alphas) for the scales ranged from .70 to .93. These factors loadings within the underlying construct explained 60.3% of the total variance.⁷² Pattern coefficients are presented in Table 1.

Table 1.

Exploratory Factor Pattern Coefficients for Principal Axis Extraction and Promax Rotation of the Eight-Factor Structure (n = 399).

Item	Factor 1	Factor 2	Factor 3	Factor 4	Factor 5	Factor 6	Factor 7	Factor 8	h ²
I want to work against injustice or corruption.	.87	−.26	−.02	−.02	.12	.04	.04	.02	.66
I want to inform the public about political events and consequences.	.84	.09	−.03	.01	.00	−.16	−.10	.12	.70
I want to hold public officials accountable.	.83	−.06	−.14	.09	.04	.03	−.11	−.02	.67
I want to point people toward possible solutions to society’s problems.	.79	−.04	.04	−.06	−.01	.02	−.02	−.00	.68
I want to motivate citizens to participate in decision making on public issues.	.77	.04	.06	−.01	−.06	.01	−.06	−.00	.67
I want to increase public awareness of significant events or breaking news.	.74	−.08	.09	.01	−.04	.15	.03	−.08	.69
I want to explain complicated issues.	.73	.11	−.02	.02	−.10	.00	.10	−.13	.66
I want to investigate the powerful.	.69	.00	.11	.02	.01	−.04	−.01	.01	.60
I enjoy following politics.	.65	.21	−.21	.08	−.05	−.21	−.08	.11	.60
I want to be a voice for underprivileged individuals or underrepresented groups.	.64	−.03	.10	−.11	.05	.09	.11	.02	.65
I want to encourage tolerance of nonmainstream groups.	.63	−.08	.00	−.15	.13	.07	.05	−.02	.61
I enjoy discussing or debating current affairs.	.59	.23	−.10	.10	.01	−.14	.04	.05	.53
I want to be able to influence the way people perceive the news industry.	.48	.10	.19	−.04	−.11	.10	.06	−.01	.50
I enjoy leading or organizing a project	−.11	.71	.04	−.06	.02	−.08	−.07	−.04	.50
I enjoy asking questions.	.09	.64	−.04	.04	−.02	.06	.068	−.09	.55
I enjoy creating or developing projects.	−.07	.58	−.06	−.09	.23	.05	−.00	.08	.53
I enjoy exerting creative control over a project.	−.05	.55	.02	−.19	.20	.01	−.07	.10	.53
I enjoy sharing my expertise with the public.	.04	.52	.20	.01	−.16	.11	.12	−.04	.52
I enjoy informing people of what they might not have known before.	.04	.49	.03	.03	−.09	.13	.08	−.02	.43
I want a job that commands a high degree of personal responsibility.	.05	.48	.16	−.03	−.01	.13	−.15	−.06	.49
I enjoy frequent interaction with new people.	.05	.47	.11	.13	.05	.00	.13	.03	.52
I enjoy learning how things work.	.07	.47	−.04	.10	.19	.00	.03	−.08	.44
I enjoy telling stories.	.13	.44	−.13	.02	−.09	.27	−.06	.09	.49
I want to be a local or national celebrity.	.01	.01	.84	−.04	−.02	−.27	−.00	.01	.71
I want to have high status in my community.	.04	.12	.78	−.10	−.11	−.14	−.00	−.02	.60
I want special access to events.	.02	−.04	.73	.05	−.03	.03	.01	.15	.58
I want opportunities to meet celebrities.	−.10	−.02	.67	.12	.10	.07	−.08	.12	.59
I want to be on TV.	.04	−.01	.64	.17	.00	−.16	−.05	−.13	.62
I want to (or expect to) earn a particular income.	.04	.04	.46	−.08	.03	.02	−.30	.01	.50
I enjoy public recognition for my work.	−.07	.29	.43	−.09	−.05	.04	−.00	−.01	.39
I enjoy sports media.	.03	−.05	−.07	.97	.04	.05	−.01	−.01	.86
I want to work in sports media.	−.02	.02	−.05	.91	−.02	−.03	.05	−.03	.76
I want to meet local, national, or international sports athletes.	−.04	.00	.20	.83	.03	.07	.01	.07	.45
I enjoyed playing sports in high school.	−.01	−.02	.01	.63	.02	.01	.02	.01	.76
I enjoy photography.	−.04	.02	−.03	.04	.81	−.04	−.00	.05	.58
I want to refine my photography skills.	.02	−.05	−.03	.00	.80	.09	−.05	−.02	.58
I enjoy editing photography or video.	.03	.19	−.05	.03	.65	−.22	−.02	−.02	.47
I want to refine my writing skills.	.00	.10	−.09	.08	.02	.69	−.05	−.01	.51
I want to work with new forms of writing.	−.08	.15	−.06	.01	.02	.74	−.04	−.01	.55
I enjoy writing.	.04	.09	−.25	.01	−.16	.67	−.11	.08	.53
I want a career that has a routine schedule (reversed).	.00	−.02	−.09	.04	.00	−.04	.77	.04	.59
I want to work regular hours (reversed)	−.05	.05	−.15	.04	−.08	−.17	.82	.02	.59
I want to have a varied (nonroutine) daily lifestyle.	.07	.09	.02	−.04	.34	.07	.38	−.03	.49
I want to travel for my job.	.07	−.02	.24	−.09	.14	.25	.33	−.03	.45
I want to avoid a major that emphasizes math.	.09	−.02	.10	.10	.10	.03	−.09	.72	.48
I want a job that allows me to use my mathematical skills. (reversed)	−.04	.07	−.09	−.05	−.13	−.05	.21	.68	.60
I enjoy working with numbers and statistics. (reversed)	.00	−.12	.01	−.21	.00	.01	.03	.58	.55
I want to avoid a major that emphasizes science.	−.03	−.04	.28	.15	−.01	.11	−.01	.44	.43
Eigenvalues	10.447	4.894	3.470	2.652	2.158	2.131	1.654	1.378
a	.93	.83	.84	.90	.80	.77	.70	.70

Note. Salient pattern coefficients are noted in bold. Figures were reduced to two decimal points. h² = communalities. Communalities are considered high if above .80. However, the more common range in the social sciences ranges from .40 to .70.⁷³

The next stage involved reducing the number of items to create a more practical and usable measure of motivations. The strong intercorrelation among the items within each factor suggested high redundancy among these items, and thus, reducing the number of items would make the scale more manageable while preserving the eight-factor structure. Both objective and subjective criteria were used to evaluate the items within each factor. Objective measures included examining the factor loadings and interitem correlations of items within each factor. Subjective analysis was used to select items that were most strongly related to the latent variables, as well as eliminating those items that had strong correlations but weak conceptual relationships to the underlying factor. Based on these criteria and theoretical logic, three to four items were selected for each factor because scale experts suggest that each factor should include a minimum of three factors, while four or five items per dimension is ideal.⁷⁴

The results, reported in Table 2, confirm the structure of the twenty-seven-item scale, with all items loading at .41 or above on its designated factor and no item loading above .32 on another factor with a cumulative variance of 66.8%.⁷⁵ Streiner argued that factors should explain at least 50% of the total variance.⁷⁶ The eight factors were social responsibility, reporting, social prestige, sports media, photography, writing, varied career, and science and numbers anxiety. The labels for each factor were created based on feedback from academics. The scholars were shown the item groupings for each subscale.

Table 2.

Exploratory Factor Pattern Coefficients for Principal Axis Extraction and Promax Rotation of the Eight-Factor Structure (n = 399).

Item	Factor 1	Factor 2	Factor 3	Factor 4	Factor 5	Factor 6	Factor 7	Factor 8
Social responsibility
I want to work against injustice or corruption.	−.05	.89	.05	.03	.04	−.19	−.03	.09
I want to point people toward possible solutions to society’s problems.	−.02	.78	.00	−.05	−.02	.09	.00	−.03
I want to inform the public about political events and consequences.	.02	.72	−.02	−.01	.02	.04	.02	−.07
I want to be a voice for underprivileged or underrepresented groups.	.03	.67	−.07	.00	−.02	.14	.05	.04
Reporting skills
I enjoy asking questions.	−.03	−.01	.00	.04	−.02	.80	−.06	−.08
I enjoy frequent interaction with new people.	.06	.04	.13	.04	−.12	.58	.04	.09
I enjoy informing people of what they might not have known before.	.05	−.03	.01	−.05	.08	.53	−.02	.06
I enjoy telling stories.	−.09	.08	−.04	−.03	.26	.49	.06	−.13
Social prestige
I want to be a local or national celebrity.	.82	.05	−.03	−.02	−.16	−.08	−.03	−.03
I want to have high status in my community.	.81	.03	−.10	−.10	−.05	.03	−.08	−.02
I want special access to events.	.73	.00	.03	−.06	.02	−.01	.11	.03
I enjoy public recognition for my work.	.47	−.05	−.10	.02	.10	.16	−.04	−.03
Sports media
I want to meet local, national, or international sports athletes.	.11	−.01	.89	.01	.05	.04	.04	−.01
I want to work in sports media.	−.10	−.03	.86	−.03	−.00	.01	−.04	−.01
I enjoyed playing sports in high school.	−.08	.04	.66	−.03	−.06	.03	−.00	.05
Photography
I enjoy photography.	−.05	−.07	−.01	.87	−.05	.01	.07	−.01
I want to refine my photography skills.	.01	.01	−.03	.83	.13	−.09	−.03	−.01
I enjoy editing photography or video.	−.04	.04	−.01	.63	−.16	.11	−.03	−.01
Writing
I enjoy writing.	−.13	.03	−.06	−.09	.78	−.05	.05	−.09
I want to work with new forms of writing.	.08	−.06	−.01	.03	.76	.03	−.04	.07
I want to refine my writing skills.	.01	.04	.07	.01	.69	.01	−.04	.08
Varied career
I want a career that has a routine schedule (reversed).	−.14	−.02	.04	−.11	−.03	−.07	.05	.76
I want to have a varied (nonroutine) daily lifestyle.	.02	.05	−.00	.20	.03	.06	−.05	.59
I want to travel for my job.	.21	.06	−.05	.03	.12	.09	−.02	.41
Science and numbers anxiety
I want to avoid a major that emphasizes math.	−.00	.10	−.01	.08	−.01	−.06	.89	−.07
I want to avoid a major that emphasizes science.	.19	.00	.10	−.03	.03	−.02	.51	.02
I want a job that allows me to use my mathematical skills (reversed).	−.14	−.17	−.16	−.10	−.04	.10	.47	.21
Eigenvalues	4.52	3.12	2.07	1.86	1.76	1.48	1.25	0.99
a	.85	.72	.78	.83	.78	.67	.79	.68
	5.408	3.490	2.208	2.010	1.845	1.459	1.255	1.075

Note. Salient pattern coefficients are noted in bold. The figures were reduced to two decimal points.

The goal is unidimensionality rather than reliability in scale development, which means it is preferred to have reliabilities that are moderate in magnitude. Thus, fewer items are preferred if they possess acceptable internal reliabilities. The unidimensionality of scale indicates that items assess a single underlying factor or scale, whereas internal consistency refers to the overall level to which items are intercorrelated. The ideal level in scale development is .80, but levels as low as .50 have been considered adequate.⁷⁷ To develop a scale, parsimony must be weighed against completeness.⁷⁸ Items that do little to explain variance only add to the length of the scale, and thus, it would not be appropriate to include all items related to motivations for a journalism degree. The scale should not be expected to cover all motivations for a journalism degree, even though other motivations may exist.

To confirm the factor structure, the twenty-seven-item scale was subjected to CFA using Amos 21.0 to assess the validity by testing the hypothesized model on a second dataset of journalism majors. Fit indices approached these levels: χ² (83, N = 310) = 667.434, p < .001; χ²/df = 2.262; CFI = .86; TLI = .83; RMSEA = .064 (with range of .057 to .070). An examination of modification indices suggested that fit could be improved if several errors were free to correlate. Specifically, a better fit would be obtained if the errors between items 22 and 23 and items 26 and 27 were correlated within each dimension. When these changes were implemented, fit was improved. Overall, the fit indices suggested a suitable fit: χ² (84, N = 310) = 563.612, p < .001; χ²/df = 1.917; CFI = .90; TLI = .88; RMSEA = .054 (with range of .048 to .061). The TLI suggested a marginally acceptable fit because it was slightly below the .90 threshold. The CFI and RMSEA suggested an acceptable fit because the CFI was at the .90 level, and the RMSEA was classified as a fair fit because it was between .05 and .08.⁷⁹ The internal consistency reliabilities from this sample were as follows: photography, α = .84; social responsibility, α = .82; sports, α = .82; writing, α = .81; social prestige, α = .80; reporting, α = .67; science and numbers anxiety, α = .53; and varied career, α = .43. Thus, the science and numbers anxiety and varied career dimensions need to be further addressed by testing these items on different samples. It is also advised to create additional items to measure these two subscales to determine the structural validity of these dimensions in future research. Reliabilities for the other half of the sample were much better for varied career (α = .79) and science and numbers anxiety (α = .68), as shown in Table 2.

Discussion

This study examined the psychometric properties of JDM. The goal was to be of service to academics by detailing the scale development and reporting process. The JDM scale was theoretically derived from the literature, qualitative research, and quantitative research. EFA was used to investigate the underlying dimensions of reasons for seeking a journalism degree, and CFA confirmed the eight-factor structure (social responsibility, reporting skills, social prestige, sports media, photography, writing, varied career, and science and numbers anxiety) on a different dataset. Table 2 shows the items comprising each factor. Each factor is overviewed to provide additional insight.

Social Responsibility

Students see journalism graduates as people who serve the public through awareness and holding public officials accountable. These items reflect a public duty to society. It is commonly believed that journalism functions to protect individuals’ rights and promote democracy. Previous research has shown this factor is a global attraction to the field.⁸⁰ Historically, journalists have perceived the news media’s duty is to investigate government practices.⁸¹

Reporting Skills

One of the most interesting observations from this research was how items related to basic reporting skills emerged as a factor, grouping “telling stories,” “asking questions,” “informing people,” and being “a voice for the underprivileged and underrepresented groups.” Further research might help indicate whether the interrelationships are a natural outgrowth of the practice of journalism or whether they might be influenced by media depictions of journalism.

Social Prestige

Journalism is often seen as a path to a higher social status and fame. It is difficult to obtain work in both journalism and acting. However, the study’s focus group research revealed students perceive journalism as a more stable and accepted career in comparison with acting. The idea of notoriety lures people to the acting craft despite its challenges.⁸²

Sports Media

Sports media training tends to take place within journalism programs, and this factor reflects this motivation for a journalism degree. Little research has assessed the characteristics, motivations, and values of sports journalists. It is known that sports journalism is one of the most popular specializations in the field of journalism especially among male students.⁸³ Based on the focus group research, some students enjoyed playing sports in high school and wanted to continue their passion as a member of the media.

Photography

Journalism is a field that tends to attract artistic types of people.⁸⁴ The focus group research showed people interested in photography select journalism as a major. Journalism students see the major as a path to develop such skills.⁸⁵ Mobile and digital technologies has enabled the visual saturation of society. However, student photographers seek a college education to learn techniques that separate them from the masses. Students are drawn into photography because of fashion, advertising, and the physical experience of working on location.⁸⁶

Writing

Previous research has found writing to be a primary reason for seeking a journalism degree.⁸⁷ Many working journalists have said a passion for writing influenced their decision to work in journalism.⁸⁸ Prior to World War II, most people interested in journalism studied English.⁸⁹

Varied Career

This factor demonstrates that journalism majors perceive it as a stimulating career path.⁹⁰ Most journalists are thought to spend their time outside of the newsroom. Thus, the perception is that they will work in a career with much excitement.

Science and Numbers Anxiety

The survey included items reflecting numbers and science apprehension, which was heard in focus groups. The fear of numbers has been found to be a common characteristic of journalism students, and majors such as journalism are not perceived as a major that substantially requires those skills.⁹¹

Conclusion

The scale efforts were intended to assist educators with advising and curricular decisions by creating a useful instrument, build theory at the operational level by providing clarity of a variable, and address practices in scale development and reporting. The JDM scale represents reasons students say they choose journalism. Overall, most motivations appear to be intrinsic in which they seek to communicate, to creatively express themselves, and to create change for public betterment.

The scale has a variety of potential applications. Administrators, educators, and advisors could survey students’ motivations to assess their interest in journalism and to what degree curriculum and classes reflect those motivations. They could also use it to help make admission choices or to communicate to internal or external stakeholders to what extent their program is representing students’ interests. Also, it would be of interest to see how newsroom organizations, internships, and academic advancement influence motivations or how external salient referents such as parents, advisors, or friends affect their motivation perceptions. In addition, the JDM scale could be used to compare students’ motivations across tracks or medium, geographic regions, public and private institutions, country, personality types, values, ethics, and so on. It could be used as an independent variable to predict self-concept, technology acceptance, learning styles, literacy levels, career goals, and major perceptions.

It is important to continue testing the scale on other samples because of the inductive reasoning involved in the scale creation process.⁹² The available sampling frame limits our ability to collect a sample that could be generalized to all undergraduate journalism majors in the world, and thus, replication is necessary on different student populations. It is important to keep in mind that self-report data reflect perceptions at one point in time. Thus, future research should determine how motivation perceptions relate to actual behaviors such as career choice. When selecting a sample to represent a larger population, it is important that samples possess proportionate distribution of demographics, backgrounds, and traits to ensure that the results can be generalized to the larger population. For creation of the scale, the inclusion of students from five different U.S. schools helped to ensure a more generalizable set of results than just including students from one school, while providing enough variance on individual scale items and subscales to establish validity of the JDM scale.⁹³

Awareness, doctoral student training, time, and journal culture are several potential reasons for the persistence of issues related to scale development. A theoretical obstacle to scientific progress is that researchers often do not communicate a justification behind their operational measures. Thus, it is difficult to dissect and replicate research based on the information provided. It is problematic to make claims about knowledge if we are creating a foundation of research on different measures measuring the same construct. Scale development can be a time-consuming and challenging process. The detail provided in this study should help guide researchers through the various stages of scale development. Communication journals could allow space to dissect measures to encourage scholars to learn best practices. Journals are an important place to have this dialogue because psychometrics is not likely extensively covered in most doctoral programs within this scholarly field. Few scales exist in journalism literature. The hope is from this review of reporting practices that a more consistent framework will be used in scale development because it is essential to ensure our foundation of knowledge is being built on sound measures.

Footnotes

Acknowledgements

We thank the National Academic Advising Association (NACADA): The Global Community for Academic Advising () for funding the project and Dr. Tim Brown from the University of Central Florida for his assistance with the project.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the National Academic Advising Association (NACADA): The Global Community for Academic Advising.

Notes

Author Biographies

Serena Carpenter is an assistant professor in the School of Journalism at Michigan State University.

August E. Grant is the J. Rion McKissick professor of Journalism in the School of Journalism and Mass Communications at the University of South Carolina.

Anne Hoag is an associate professor in the Telecommunications Department at Penn State University–University Park.