Next Article in Journal
Volatiles Emitted by Three Genovese Basil Cultivars in Different Growing Systems and Successive Harvests
Previous Article in Journal
Runoff for Russia (RFR v1.0): The Large-Sample Dataset of Simulated Runoff and Its Characteristics
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Data
Volume 8
Issue 2
10.3390/data8020032
data-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Review

LGCM and PLS-SEM in Panel Survey Data: A Systematic Review and Bibliometric Analysis

by
Zulkifli Mohd Ghazali
1,
Wan Fairos Wan Yaacob
2,3,* and
Wan Marhaini Wan Omar
4
1
Mathematical Sciences Studies, College of Computing, Informatics and Media, Universiti Teknologi MARA, Cawangan Perak, Kampus Tapah, Tapah Road 35400, Perak, Malaysia
2
Mathematical Sciences Studies, College of Computing, Informatics and Media, Universiti Teknologi MARA Cawangan Kelantan, Kampus Kota Bharu, Kota Bharu 15050, Kelantan, Malaysia
3
Institute for Big Data Analytics and Artificial Intelligence (IBDAAI), Kompleks Al-Khawarizmi, Universiti Teknologi MARA, Shah Alam 40450, Selangor, Malaysia
4
Faculty of Business and Management, Universiti Teknologi MARA Cawangan Kelantan, Kampus Kota Bharu, Kota Bharu 15050, Kelantan, Malaysia
*
Author to whom correspondence should be addressed.
Data 2023, 8(2), 32; https://doi.org/10.3390/data8020032
Submission received: 13 December 2022 / Revised: 19 January 2023 / Accepted: 22 January 2023 / Published: 30 January 2023
(This article belongs to the Topic Recent Advances and Applications in Partial Least Squares Structural Equation Modeling (PLS-SEM))
Browse Figures
Versions Notes

Abstract

:
The application of Latent Growth Curve Model (LGCM) and Partial Least Square Structural Equation Modeling (PLS-SEM) has gained much attention in panel survey studies. This study explores the distributions and trends of LGCM, and PLS-SEM used in panel survey data. It highlights the gaps in the current and existing approaches of PLS-SEM practiced by researchers in analyzing panel survey data. The integrated bibliometric analysis and systematic review were employed in this study. Based on the reviewed articles, the LGCM and PLS-SEM showed an increasing trend of publication in the panel survey data. Though the popularity of LGCM was more outstanding than PLS-SEM for the panel survey data, LGCM has several limitations such as statistical assumptions, reliable sample size, number of repeated measures, and missing data. This systematic review identified five different approaches of PLS-SEM in analyzing the panel survey data namely pre- and post-approach with different constructs, a path comparison approach, a cross-lagged approach, pre- and post-approach with the same constructs, and an evaluation approach practiced by researchers. None of the previous approaches used can establish one structural model to represent the whole changes in the repeated measure. Thus, the findings of this paper could help researchers choose a more appropriate approach to analyzing panel survey data.

1. Introduction

Over the past few decades, various survey studies have been conducted using different types of survey designs. Many of them used cross-sectional survey design that is able to measure variation in the individuals of a population [1,2,3,4] at one point in time. However, in recent years, the development from cross-sectional to panel survey studies can be seen to escalate [5,6,7] in longitudinal studies. The panel survey data has been used widely in several areas such as education, medicine, psychology, behavior, and many more [8,9,10]. This type of study allows the researcher to measure variation at the individual level repeatedly on the same sample of units at different points of time. Through panel survey data, the trend and factors influencing those changes can also be observed.
From a methodological perspective, there are several methods that can be used to analyze cross-sectional survey data types. The most commonly used methods are based on Structural Equation Modelling (SEM). SEM offers two methods which are (i) covariance-based SEM (known as CB-SEM), and (ii) variance-based SEM (known as Partial Least Square (PLS-SEM)) method. These methods are often used to identify multiple statistical relationships simultaneously through visualization and model validation. It is more suitable for complex models compared to the traditional method such as Multiple Linear Regression and Analysis of Variance (ANOVA). CB-SEM and PLS-SEM have their own strengths and weaknesses depending on the data structures and assumptions of the methods.
While in the panel survey data, a method based on the CB-SEM framework known as Latent Growth Curve Model (LGCM) is commonly being used compared to the PLS-SEM. The LGCM has gained its popularity largely in behavioral sciences research. This method is widely used in many areas including social behavioral research, psychology, clinical, developmental, educational research, learning and memory, and personality [11,12,13,14,15,16]. The LGCM has the advantage of analyzing the developmental trajectory of a single person and capturing individual variations over time. This means the method can assess the changes in intra-individual (within the individual) as well as inter-individual (between individuals) variation. It can also identify the important predictor variables that contribute to the individual’s growth change over time. Although LGCM listed several advantages, this method can still be improved as highlighted by [11,17,18]. Among the issues concerned are statistical assumptions, reliable sample size, number of repeated measures, and missing data. Despite its limitation, this method is still the choice of researchers for analyzing panel survey data compared to the PLS-SEM.
On the other hand, PLS-SEM only gained popularity in analyzing cross-sectional survey data but not in panel survey data. In a cross-sectional survey study, the PLS-SEM showed good performance in handling non-normal data and small sample size. According to [9], PLS-SEM showed higher robustness in situations of non-normal data and small sample size. It also shows a better result with a small sample size when a model has many constructs and a large number of items [19,20,21]. However, the use of PLS-SEM in panel survey data seems to be seemingly underrated as cross-sectional survey data, even though the PLS-SEM is good in handling several highlighted issues such as statistical assumptions and reliable sample size in LGCM. The previous approaches for panel survey data using PLS-SEM were unable to establish one structural model to represent the whole changes in the repeated measure. Current approaches also cannot capture the individual trajectory, mean of the trajectory of the sample or entire group, the evaluation of individual differences in trajectories, and assess the potential incorporation of predictors of individual differences in trajectories. A review of the previous approaches of these path analyses in longitudinal studies does not consider systematic literature review methodology [22]. Thus, there is a need to review the current and existing PLS-SEM approaches using SLR and Bibliometric analysis for panel survey data in identifying the gap in existing methods for improvement.
Hence, this study aimed to explore the distributions and trends of LGCM, and PLS-SEM used in panel survey data. It highlights the gaps in the current and existing approaches of PLS-SEM practiced by researchers in analyzing panel survey data. It focuses on answering the following research questions; (i) What is the distribution and trend of LGCM and PLS-SEM in a panel survey study? (ii) What are the reasons for the lack of application of PLS-SEM in panel survey data? and (iii) What is the existing framework or procedure of PLS-SEM in analyzing the panel survey data? This study employs the integrated bibliometric analysis and systematic review because the way of reviewing the existing literature is more systematic, and more comprehensive compared to the classical literature review [23,24,25]. Through a systematic review, further investigation and identification of the reasons for the lack of application for PLS-SEM in panel survey data can be discovered. Exploration of the existing framework or procedure of PLS-SEM could help the researcher to identify the method for improvement in analyzing the panel survey data.

2. Related Work

2.1. Panel Survey Data

A panel survey is a type of survey method that involves the process of gathering data from the same sample over a period of time. It is one of the longitudinal study types that is conducted over an extended period of time. The data collected from this panel survey are referred to as panel survey data. Panel survey data are commonly used to measure the behavior of people over time including their thoughts, attitudes, feelings, emotions, and many more [26,27,28]. It can measure the changes in behavior over time and examine the factors that influence that change. In the context of statistical methods, the LGCM and PLS-SEM are two methods that are used for analyzing the panel survey data. These two methods can handle this type of data since both can assess the measurement model (reliability and validity) and structural model. This is because these methods used Structural Equation Modeling as a basis of the framework and followed its criteria.

2.2. Latent Growth Curve Model

The latent Growth Curve Model (LGCM) created by [29,30] has grown to be a better method for addressing issues about individual behavior change and assessing the factors that contributed to the change simultaneously. The LGCM is a combination of the growth curve model (GCM) and structural equation modeling (CB-SEM). According to [31], the LGCM is a special case of confirmatory factor analysis (CFA) in CB-SEM and followed all underlying assumptions of CB-SEM. The growth of LGCM has become more popular in panel survey study since it can measure the changes in individuals and groups (known as trajectory) over time. Furthermore, it can also assess the factors that influence the trajectory.

2.3. PLS-SEM for Panel Survey Data

The PLS path modeling or PLS-SEM was created by [32,33] and some extensions were suggested by [34]. Over the last few decades, there have been numerous introductory articles on this methodology (e.g., [35,36,37,38]). However, in the panel survey studies, the application of this method is very limited compared to the cross-sectional studies [22]. This is because the exploration and the procedure of PLS-SEM for analyzing panel survey data are not consistent since it was used differently by the authors in several research articles [39,40,41,42,43].

3. Materials and Methods

This section explains the methodology used in this study. This study used an integrated systematic literature review (SLR) and bibliometric analysis for the review process [24,25,44].

3.1. Phase 1—Systematic Literature Review (SLR)

In the systematic review, the process of reviewing followed the review protocol, publication standard, or established guideline. The review protocol is equivalent to a research design in social sciences research. It is very important to decide which review protocol, publication standard, or established guideline is to be used at the beginning of the study [45]. This study adapted the established guideline by [46,47]. This established guideline was developed specifically for the education field. However, the guideline is suitable to be adapted in other fields, and it has been used in many other fields too. Based on this established guideline, this study started with the formulation of the research problems, followed by a systematic searching strategy (identification, screening inclusion, eligibility, and quality appraisal), data extracting, analyzing, and synthesizing (theme generation).

3.1.1. Formulating the Research Problems

The formulation of the research problems or the research questions for this study is based on the PICo [48,49]. PICo is used as a guideline to develop the research questions. PICo consists of three main concepts which are population or problem, interest, and context. In this study, the population can be described as panel survey data with several interests such as distributions and trends, limitations, and procedures. Then, this study described the context of statistical methods such as LGCM and PLS-SEM. Based on this concept, these research questions were created: “what are the distributions and trends of LGCM and PLS-SEM in a panel survey study?”, “what are the limitations of PLS-SEM in a panel survey data?” and “what is the existing framework or procedure of PLS-SEM in analyzing the panel survey data?”.

3.1.2. Systematic Searching Strategies

In this stage, there are three main processes of searching strategies: (i) searching the literature (identification), (ii) screening the inclusion, and (iii) eligibility.
  • Searching the Literature (Identification)
Web of Science Core Collection (WoSCC) and Scopus are two bibliographic databases that are often regarded as the most comprehensive data sources for a variety of uses [50]. WoSCC was established around 2014 and previously known as the Web of Science (WoS) [51]. It was the first comprehensive international bibliographic database produced by Thomson Reuters in 1997. WoSCC consists of ten sub-databases and this study used eight sub-databases from the year 1992 to 2022. Among the sub-databases are Social Sciences Citation Index (SSCI), Science Citation Index Expanded (SCI-EXPANDED), Emerging Sources Citation Index (ESCI), Conference Proceedings Citation Index – Social Science & Humanities (CPCI-SSH), Arts & Humanities Citation Index (A&HCI), Conference Proceedings Citation Index – Science (CPCI-S), Book Citation Index – Social Sciences & Humanities (BKCI-SSH), and Book Citation Index – Science (BKCI-S). As a result, it eventually rose as the top choice of bibliographic database for bibliometric analyses, research appraisal, journal selection, and other duties [52]. In 2004, Elsevier introduced Scopus and established a solid reputation for dependability and earned a spot-on level with other comprehensive bibliographic databases over time [50,53]. Apparently, Scopus has a wider coverage, and thus it is useful for mapping a smaller research field as in the emerging topic of this study [54]. WoSCC and Scopus are also multidisciplinary and selective databases that are composed of a variety of specialized indexes, grouped according to the type of indexed content or by theme, [55]. Hence, both databases were employed as the bibliographic database for this study particularly to search for the right literature. For that reason, keywords are required to create the search string. In this study, the keywords were derived from the developed research questions as suggested by [56] as shown in Table 1. Based on this search string, a total of 3850 articles were retrieved automatically from the Scopus and WoSCC bibliographic databases. In addition, the stopping rule of searching the article is based on the rule of thumb as suggested by [57], where the search can stop when repeated search results are found in the same references, with no new results.
2.
Screening the Inclusion
In the screening process, the articles were refined based on five criteria in the bibliographic database: (i) timeline, (ii) language, (iii) document type, (iv) subject area, and (v) type of data. The details for each criterion are explained in Table 2. In this stage, 2640 articles met the criteria and qualified for the next process.
3.
Eligibility
The eligibility process involved the review of the title, keyword, and abstract. [58] suggested that a researcher should review the conclusion if the information in the abstract cannot give the general picture of the article. The selection of articles is based on the inclusion criteria (Table 2) contained in either the title, keyword, or abstract. After this selection process, the articles were checked for duplication according to the title and the redundant articles were removed. Hence, 1726 articles were selected after the removal process.

3.2. Phase 2—Bibliometric Analysis

The bibliometrics method was first introduced in 1969 by a scholar named Pritchard. The term bibliometric is elaborated as an information and library sciences research area which employs a quantitative approach and analyzes the bibliographical data of among others, the year of publication, country of origin, authors, etc. [59]. The bibliometric method employs quantitative analysis of empirical data published in prior literature to study the trends of publication within various research domains. Furthermore, it enables researchers to examine the body of literature in their field of study and identify the major themes [54,60]. Using bibliometric analysis allows researchers to explore the trends, reader usage, citation pattern, knowledge base, author network, and significance of the subject [61]. Bibliometric analysis is often combined with science mapping techniques to visualize the intellectual structure of a particular research field [62]. Visualization requires visual tools such as VosViewer, Gephi, or Pajek, which have been used extensively in management and science research. In this study, bibliometric analysis was employed to analyze citation-based analysis, co-word analysis or keyword co-occurrence analysis, and co-authorship analysis, which are considered the most common ones using this method.

3.2.1. Data Extraction

The process of data extraction was followed by data requirements of bibliometric analysis such as the author’s names, citations, titles, journals, DOI, references, abstracts, keywords, and author affiliations [46]. The data from each bibliographic database was extracted into an excel file and merged following the Scopus format. Then, data were exported into VOSviewer for constructing and visualizing the information. Next, the thresholds such as the minimum number of publications, citations, and occurrence of keywords were specified for analysis of science mapping.

3.2.2. Analyzing and Synthesizing the Data

The bibliometric analysis consists of two techniques which are performance analysis and science mapping. This study used performance analysis to determine the distribution and the trend of the publication related to the panel survey data. Besides that, the analysis of science mappings such as Co-authorship, Keyword Co-occurrence, Citation, and Co-citation Analysis was used to examine the relationships between the research constituents [46].

3.3. Phase 3—Content Analysis

In this phase, the procedure from SLR, which is quality appraisal, is used to select suitable articles for content analysis. The content analysis was used to generate the themes to explain the findings related to the PLS-SEM in panel survey data.

3.3.1. Quality Appraisal

In this stage, selected articles that are related to the PLS-SEM were chosen based on citation analysis in the bibliometric analysis. The total number of articles related to the PLS-SEM was 296, after the eligibility process. However, for the content analysis, only the top 100 most cited articles were included in the quality appraisal process. The quality appraisal is very important in the systematic literature review as suggested by [63]. In this process, the full articles were examined by the research team to select the most suitable articles that are related to the procedure of PLS-SEM in analyzing the panel survey data. After the quality appraisal process, 34 articles were selected for the final review (Figure 1, Table A1).

3.3.2. Theme Generation

In this stage, the theme was generated based on the 34 articles reviewed. The themes were classified into PLS-SEM approaches and their imitations, and the procedure of the approaches.

4. Results

This section discusses the distributions and trends of LGCM and PLS-SEM in panel survey data, the limitations of PLS-SEM in panel survey data, and the existing framework or procedure of PLS-SEM in analyzing the panel survey data. This discussion reflects the research questions stated in the early section.

4.1. Distributions and Trends

To answer the first research question, the discussion discovers the growth of publications, co-authorship, citation, co-citation, and co-occurrences of keywords.

4.1.1. Growth of Publications

Figure 2 shows the annual growth of publications related to the panel survey data that used the Latent Growth Curve Model (LGCM) and PLS-SEM as the main statistical methods in the analysis. These publications were retrieved from Scopus and Web of Science Core Collection (WoSCC) databases from 1986 to 2022. Based on the graph, the publications show an increasing trend from 2006 to 2022. Figure 2 also shows the annual growth of publications related to LGCM and PLS-SEM separately. Both annual growths of publications show an increasing trend from 2006 to 2022. However, LGCM is more outstanding compared to PLS-SEM as a statistical method to analyze panel survey data.
All 1726 retrieved articles were published in 638 different journals, with 2.71 articles per journal on average. Out of these 638 journals, 365 (57.21%) published only one article, 118 (18.59%) published two articles, and 115 (24.29%) published more than two articles. Table 3 shows the top ten journals contributing to the panel survey data. Based on the total citations, Development Psychology journal is the most cited journal with 4231 citations, followed by Structural Equation Modeling, and Psychology and Aging journal with 1267 and 1061 citations respectively. However, in terms of total publications, no journals show an outstanding performance since the number of publications is close to each journal.

4.1.2. Co-Authorship Analysis

The main purpose of co-authorship analysis is to examine the interactions among scholars related to the panel survey data. Based on the retrieved records, 4481 authors contributed 1726 articles in the panel survey data. Out of 4481, only 459 authors met the threshold of at least 2 publications and 25 citations. Figure 3 shows that the connection between clusters is small and only 8 clusters are connected to each other. This result indicates that the majority of productive authors are independent researchers and the cluster formed by the researchers working on the panel survey data is weak, and the scale of co-authorship cooperation is small and limited.

4.1.3. Citation Analysis

The citation analysis was used to identify the most influential publications in the research field. The purpose is to gain an understanding of the intellectual dynamics of the research field. In this analysis, the most influential articles were selected based on the highest number of total citations and analyzed according to two statistical methods which are LGCM and PLS-SEM. Table 4 and Table 5 show the lists of the top 5 most cited articles in LGCM and PLS-SEM. McArdle J.J and Epstein D. (1987) is the most cited article according to LGCM, with 653 total citations. While in PLS-SEM, Limayen M and Cheung C.M.K. (2008) is the most cited article with 369 total citations.
In the context of relationships among publications, most of the authors work independently, which indicates a weak relationship. The relationships among authors according to the citations can be seen in Figure 4. The citation analysis for 1726 articles revealed that 494 articles met the threshold of 25 minimum number of citations of the document. The network visualization map shows that only a few clusters are connected to each cluster, even though those publications have the highest number of citations such as McArdle J.J. and Epstein D. (1987).

4.1.4. Co-Citation Analysis

Co-citation analysis of cited references was performed as well. By definition, the reference can be a co-citation if the two documents are cited together by another document [63]. As shown in Figure 5, each point represents the cited author, and the color of the points is according to the number of co-citations. A total of 88,731 cited authors were detected, and only 800 authors met the threshold in which the minimum citation of an author is 800. As seen in Figure 5, 800 authors formed 7 different clusters that provide information related to the co-citation of this study. Overall, most of the co-citations are related to statistical methods such as the Latent growth curve model, evaluation in structural equation modeling, evaluation in PLS-SEM, and procedure in the simulation study. The highest total link strength in co-citation analysis is Muthen and McArdle, and the article is related to the simulation study and latent curve analysis.

4.1.5. Keyword Co-Occurrence Analysis

The co-occurrence analysis focuses on the examination of the actual content of the publications based on the words derived from the author’s keyword. This analysis can determine the trend of research topics in recent years. Figure 6 shows the network visualization map of the co-occurrence of keywords related to the panel survey data. Based on this analysis, 3875 keywords were retrieved from 1725 articles. However, there were only 49 keywords that met the minimum threshold of occurrences number of at least 10. As seen in Figure 6, 49 keywords formed 9 different clusters that provided information about the related topic of this study. The largest cluster was the red and green clusters which consisted of 11 keywords for each cluster. In addition, Table 6 shows the list of keywords as well as their co-occurrence frequencies in each cluster. In the context of the research topic, longitudinal study and adolescence showed the highest co-occurrences in this study with 292 and 169 repeated keywords, respectively. While in the context of statistical method, the Latent growth curve model was the most used in the analysis, with 294 co-occurrences keywords, followed by structural equation modeling, PLS-SEM, and partial least squares with 22, 14, and 11 respectively. Besides, the keywords in the same cluster shared a similar topic. Generally, most of the research topic for each cluster is related to mental health, psychology, child and adolescent development, and lifestyle.
To examine the trend of research topics in the recent year, an overlay visualization map was produced. Figure 7 shows the co-occurrences of the keywords according to the time (in years). Based on the overlay visualization map, there were a few research topics in recent years, such as mental health, life satisfaction, cognition, aggression, effortful control, children, social media, and COVID-19. In the context of statistical methods, PLS-SEM has been used in recent years to analyze the panel survey data.
Overall, bibliometric analysis has fulfilled the first research question related to the trend of the Latent growth curve model (LGCM) and PLS-SEM in panel survey data. The result shows that the distribution, trend, and application of LGCM are more outstanding than PLS-SEM in analyzing the panel survey data. To answer the next two research questions, content analysis was employed. Content analysis is focused on exploring the PLS-SEM in analyzing the panel survey data. The reason behind exploring the PLS-SEM in the panel survey data is due to the bibliometric analysis that shows the development of this method is not well developed, even though this method has good potential in handling panel survey data.

4.2. Themes Generation

To answer the other two research questions related to the limitations of PLS-SEM in panel survey data and the procedure of the existing approach of PLS-SEM in panel survey data, content analysis was employed. This section explains the details of content analysis on the PLS-SEM approach for panel survey data based on the selected top 100 most cited papers. The explanation of the content analysis is divided into two different themes: (i) identification of PLS-SEM approaches and their limitations, and (ii) procedures of the method.

4.2.1. Identification of PLS-SEM Approaches and Their Limitations

The exploration of the PLS-SEM approach is explained according to the evaluation of the measurement, and the structural model. Based on the reviewed articles, most of the researchers used the standard procedure to evaluate the measurement model as suggested by [64]. The measurement model involved the evaluation of indicator reliability, internal consistency reliability, convergent validity, and discriminant validity.
However, when evaluating the structural model, most of the top researchers use different approaches and procedures. As a summary, five approaches and procedures are identified to be used by researchers. The first approach is suitable for two periods of time and uses the different latent constructs at the pre-evaluation and post-evaluation named pre- and post-approaches with different constructs. The main purpose of this approach is to evaluate relationships between the exogenous variable at pre-evaluation, and the endogenous variable or outcome at post-evaluation. The relationship was evaluated using partial least squares (PLS) in the structural model. The second approach is known as the Path Comparison approach, and it is suitable for two periods of time. This approach uses the same latent construct at the initial (t0) and end (t1) of the evolution. This approach can measure the relationship of the latent construct using path analysis and the impact of time on the PLS model using a t-test. The third approach is named the Cross-Lagged Panel Method (CLPM), and the main purpose is to measure the direction, strength, and cause-effect relationship among latent constructs over time. This method is also suitable for two periods of time and uses the same construct at time 1 and time 2. The fourth approach also involves two periods of time and measures the same latent constructs at pre-evaluation and post-evaluation named pre- and post-approaches with the same constructs. The difference between this method compared to the other three is regarding the latent score used to develop the PLS model. This method uses the differences in scores from time 1 and time 2 to develop the PLS model. The fifth approach is the evaluation approach which involves more than two waves of time and uses the same constructs for evaluating participants over time. The main purpose is to evaluate the direct and indirect effects over time. This approach uses path analysis to evaluate the direct effects between latent constructs and indirect effects over time using a bias-corrected confidence interval.
Though there are five different approaches of PLS-SEM in analyzing the panel survey data, these approaches still have limitations and spaces for improvement. The obvious limitation for models 1, 2, 3, and 4 is related to the number of waves for the study, since these approaches are only suitable for two periods of time. In addition, these four models focus on the pre- and post-evaluations and do not measure the evaluations of effects over time. Besides that, model 5 has the limitation in evaluating the growth of trajectory even though this method is capable of handling studies with more than two periods of time. This model does not have one structural model to represent the whole changes in the repeated measure. In addition, this model cannot capture an individual trajectory, the mean of the trajectory of the sample or entire group, the evaluation of individual differences in trajectories, and assess the potential incorporation of predictors of individual differences in trajectories. Furthermore, this model is not flexible to handle the latent constructs simultaneously as independent and dependent in the same model, allowing for complex representations of growth and correlations of change. Table 7 shows the summary of the five different approaches practiced by researchers in analyzing panel survey data.

4.2.2. Procedure of the Approaches

This section explains the details of the procedure for several PLS-SEM approaches practiced by the researchers in analyzing the panel survey data. The procedure of the approach is explained according to the data collection phase and analysis phase. Table 8 shows the summary of the procedure for five approaches in PLS-SEM to analyze the panel survey data.
  • Model 1: Pre and Post Approach with Different Construct
This approach involves two phases of time in the data collection procedure. At time 1, the participants will be evaluated using the first set of questionnaires that consist of exogenous variables. The different sets of questionnaires that consist of the endogenous variable will be used for the second evaluation. In the analysis phase, the measurement and structural model will be evaluated. All the latent constructs will be evaluated based on reliability and validity. For the structural model, one PLS model will be established together with the path coefficients to perform the bootstrap resampling procedure to examine the significance of the paths.
  • Model 2: Path Comparison Approach
For this approach, the participants will be evaluated two times with the same questionnaire. In the measurement model, the reliability and validity for each construct at time 1 and time 2 will be developed separately. For the structural model, two PLS models will be developed separately according to the time (time 1 and time 2). Then, calculate the t-test using the formula suggested by [72] for comparing the corresponding path coefficient in both models. This analysis will examine the strength of the relationship between the paths over time.
  • Model 3: Cross-Lagged Approach
The procedure of this approach for the data collection phase and the measurement model evaluation is the same as Model 2. For the structural model, the analysis starts with the mean score comparison between latent construct time 1 and time 2. The purpose is to determine whether the mean score of the latent construct at time 2 will be higher than at time 1. Next, to determine the cause-effect relationship between latent constructs, the cross-lagged panel model will be employed.
  • Model 4: Pre and Post Approach with the Same Construct
The procedure of this approach for the data collection phase and measurement model evaluation is also the same as Models 2 and 3. For the structural model, the analysis starts with the comparison of the indicator of latent construct between time 1 and time 2 using paired t-test. If the result of the paired t-test has significant differences, then the new indicators are computed based on the differences between indicators at time 1 and time 2. Next, one PLS model will be developed based on the new indicators to determine the effects between change constructs.
  • Model 5: Evaluation Model
The data collection phase involves more than two periods of time with the same questionnaire. For the measurement model, the constructs will be evaluated according to the time. While in the structural model, the analysis starts by developing one PLS model for each period of time. In this stage, the direct effect and carry-over effect will be examined based on the path analysis. Then, to test the changes in the path coefficient over time, the bias-corrected confidence interval is computed. Next, paired t-test of the changes in the level of the construct over time is computed.

5. Discussion

Structural Equation Modeling (SEM) is one of the flexible methods for analyzing survey data. This method is used as a statistical tool for evaluating the relation between latent and observed variables [73]. SEM can be defined as a combination of several multivariate analysis techniques [74], such as path analysis [75] and the common factor or latent variable model [76]. Thus, this study reviewed the methodology that used SEM as a base framework for analyzing panel survey data. There are two methods that have been discovered in this study which are LGCM and several approaches in the PLS-SEM. The trend of publications related to the panel survey data is increasing over the year. The findings show that the application of LGCM is preferable compared to the PLS-SEM in analyzing the panel survey data. We can see the pattern in the bibliometric analysis where the findings are dominated by the LGCM. This is because the ability and flexibility of LGCM in handling panel survey data are better than the PLS-SEM. Among the ability of LGCM, it can describe the developmental trajectory of a single person and capture individual variations over time. In other words, this method can assess the changes in intra-individual (within the individual) as well as inter-individual (between individuals) variation. LGCM can also identify the important predictor variables that contribute to the individual’s growth change over time. [77] described the several advantages of LGCM which permits the investigation of inter-individual differences in change over time and allows the researcher to investigate the antecedents and consequences of change. LGCM also provides group-level statistics such as mean growth rate and mean intercept, can test hypotheses about specific trajectories, and allows the incorporation of both time-varying and time-invariant covariates. This could be the main reason why LGCM is preferable compared to the PLS-SEM, even though it has several limitations. According to [22], the existing approaches of PLS-SEM in panel survey data still have limitations. The approaches also show a lack of flexibility in analyzing the panel survey data in one structural model. Thus, this study employed content analysis to identify the existing approaches of PLS-SEM in analyzing panel survey data and its limitations.
The findings show that there are five existing approaches of PLS-SEM that have been used in analyzing panel survey data. Among the existing approaches of PLS-SEM, the Evaluation approach is the most flexible approach in analyzing panel survey data. Thus, this study discussed this approach more than the other four approaches. This approach consists of three stages in analyzing the panel survey data. The analysis measured the direct effect and the special effect which is the carry-over-effect. Carry-over-effects are effects from one construct at one point in time to the same construct at a subsequent point in time [78]. In stage one, the direct effect and carry-over-effect are assessed by estimating the single PLS model separately across the time. With this, the separate direct effect between the endogenous and exogenous predictors across time can be assessed. Hence, one structural model to access the whole changes (trajectories) and the factors that influence those changes simultaneously cannot be established. In the second stage, the multi-group analysis is employed to assess the strength of direct effects and the carry-over-effects over time. This strength is measured by the changes in the size of the path coefficient and bias-corrected confidence interval. The limitation at this stage is that the factors that influence those changes in the carry-over-effect simultaneously in one structure modal cannot be measured. In the last stage, paired t-test is employed to assess the mean difference between the constructs. The limitation in this stage is that only the mean difference for two points at a time for each construct can be assessed. In addition, the paired t-test requires a few assumptions and the most concern for the researcher is the distributional assumption. Hence, all these stages in the Model 5 approach do not have one structural model to represent the whole changes in the repeated measure. In addition, current approaches cannot capture the individual trajectory, mean of the trajectory of the sample or entire group, the evaluation of individual differences in trajectories, and assess the potential incorporation of predictors of individual differences in trajectories. Consequently, with all these limitations, PLS-SEM is less frequently used for analyzing panel survey data.

6. Conclusions

In conclusion, this study explored the distributions and trends of publications related to the panel survey data. This study also explored the trends of publications according to the Latent Growth Curve Model (LGGM) and PLS-SEM in analyzing the panel survey data. The records were retrieved from the bibliographic databases of Scopus and Web of Science Core Collection (WoSCC). The trends of publications related to the panel survey data showed an increasing trend. However, in the context of the statistical method, the LGCM is preferable compared to the PLS-SEM in analyzing the panel survey data, even though the LGCM has several limitations as highlighted in previous studies. This is because the PLS-SEM shows a lack of capability in handling panel survey data, even though it has five different approaches in analyzing them. The most flexible approach of the PLS-SEM in handling panel survey data is model 5 since it can measure the direct effect, carry-over effect, and the changes of path coefficients over time. However, based on the review, this approach still has some space for improvement. This method cannot capture an individual trajectory, the mean of the trajectory of the sample or entire group, the evaluation of individual differences in trajectories, and assess the potential incorporation of predictors of individual differences in trajectories. Besides, these current approaches are not as flexible as LGCM since it has the ability to use variables simultaneously as independent and dependent in the same model. Therefore, this systematic review could help researchers choose a more suitable method to analyze panel survey data.

Author Contributions

Z.M.G.: Conceptualization, Data Collection, Methodology, Formal Analysis, Visualization, Preparation of original draft. W.F.W.Y.: Conceptualization, Methodology, Writing, Review & Editing. W.M.W.O.: Methodology, Review & Editing. All authors have read and agreed to the published version of the manuscript.

Funding

The authors would like to thank Universiti Teknologi MARA (UiTM) for funding this research under PYPA.

Acknowledgments

The authors would like to thank the reviewers for their helpful and constructive comments and suggestions that greatly contributed to the improvements of the final version of this paper.

Conflicts of Interest

The authors declare no conflict of interest.

Appendix A

Table
Table A1. List of Evaluated Articles in Content Analysis.
Table A1. List of Evaluated Articles in Content Analysis.
NoAuthorsTitleYearJournalDOI
1Limayem M., Cheung C.M.K.Understanding information systems continuance: The case of Internet-based learning technologies2008Information and Management10.1016/j.im.2008.02.005
2Baer J.S., Sampson P.D., Barr H.M., Connor P.D., Streissguth A.P.A 21-year longitudinal analysis of the effects of prenatal alcohol exposure on young adult drinking2003Archives of General Psychiatry10.1001/archpsyc.60.4.377
3Bontis N., Booker L.D., Serenko A.The mediating effect of organizational reputation on customer loyalty and service recommendation in the banking industry2007Management Decision10.1108/00251740710828681
4Islam A.K.M.N.Investigating e-learning system usage outcomes in the university context2013Computers and Education10.1016/j.compedu.2013.07.037
5Barnes S.J., Mattsson J., Sørensen F.Remembered experiences and revisit intentions: A longitudinal study of safari park visitors2016Tourism Management10.1016/j.tourman.2016.06.014
6Nelson B., Martin R.P., Hodge S., Havill V., Kamphaus R.Modeling the prediction of elementary school adjustment from preschool temperament1999Personality and Individual Differences10.1016/S0191-8869(98)00174-3
7Hannula-Sormunen M.M., Lehtinen E., Räsänen P.Preschool Children’s Spontaneous Focusing on Numerosity, Subitizing, and Counting Skills as Predictors of Their Mathematical Performance Seven Years Later at School2015Mathematical Thinking and Learning10.1080/10986065.2015.1016814
8Bronstein P., Ginsburg G.S., Herrera I.S.Parental predictors of motivational orientation in early adolescence: A longitudinal study2005Journal of Youth and Adolescence10.1007/s10964-005-8946-0
9Sosik J.J., Potosky D., Jung D.I.Adaptive self-regulation: Meeting others’ expectations of leadership and performance2002Journal of Social Psychology10.1080/00224540209603896
10Chen C.-P., Lai H.-M., Ho C.-Y.Why do teachers continue to use teaching blogs? the roles of perceived voluntariness and habit2015Computers and Education10.1016/j.compedu.2014.11.017
11Benitez J., Chen Y., Teo T.S.H., Ajamieh A.Evolution of the impact of e-business technology on operational competence and firm profitability: A panel data investigation2018Information and Management10.1016/j.im.2017.08.002
12Gupta V.K., Huang R., Niranjan S.A longitudinal examination of the relationship between Team Leadership and Performance2010Journal of Leadership and Organizational Studies10.1177/1548051809359184
13Palos-Sanchez P., Saura J.R., Martin-Velicia F.A study of the effects of programmatic advertising on users’ concerns about privacy overtime2019Journal of Business Research10.1016/j.jbusres.2018.10.059
14Gegenfurtner A.Dimensions of Motivation to Transfer: A Longitudinal Analysis of Their Influence on Retention, Transfer, and Attitude Change2013Vocations and Learning10.1007/s12186-012-9084-y
15Wei Y., Zhu X., Li Y., Yao T., Tao Y.Influential factors of national and regional CO2 emission in China based on combined model of DPSIR and PLS-SEM2019Journal of Cleaner Production10.1016/j.jclepro.2018.11.155
16Palos-Sanchez, P; Saura, JR; Martin-Velicia, FA study of the effects of programmatic advertising on users’ concerns about privacy overtime2019Journal Of Business Research10.1016/j.jbusres.2018.10.059
17Roemer E.A tutorial on the use of PLS path modeling in longitudinal studies2016Industrial Management and Data Systems10.1108/IMDS-07-2015-0317
18Saeed K.A., Abdinnour S., Lengnick-Hall M.L., Lengnick-Hall C.A.Examining the Impact of Pre-Implementation Expectations on Post-Implementation Use of Enterprise Systems: A Longitudinal Study2010Decision Sciences10.1111/j.1540-5915.2010.00285.x
19Roxas B.Effects of entrepreneurial knowledge on entrepreneurial intentions: A longitudinal study of selected South-east Asian business students2014Journal of Education and Work10.1080/13639080.2012.760191
20Jung D.I., Sosik J.J.Effects of group characteristics on work group performance: A longitudinal investigation1999Group Dynamics10.1037/1089-2699.3.4.279
21Courty A., Godart N., Lalanne C., Berthoz S.Alexithymia, a compounding factor for eating and social avoidance symptoms in anorexia nervosa2015Comprehensive Psychiatry10.1016/j.comppsych.2014.09.011
22Marjoribanks K.Family background, social and academic capital, and adolescents’ aspirations: A mediational analysis1997Social Psychology of Education10.1023/A:1009602307141
23Piyathasanan B., Mathies C., Patterson P.G., de Ruyter K.Continued value creation in crowdsourcing from creative process engagement2018Journal of Services Marketing10.1108/JSM-02-2017-0044
24Gray D.M., D’Alessandro S., Johnson L.W., Carter L.Inertia in services causes and consequences for switching2017Journal of Services Marketing10.1108/JSM-12-2014-0408
25Pai H.-C.An integrated model for the effects of self-reflection and clinical experiential learning on clinical nursing performance in nursing students: A longitudinal study2016Nurse Education Today10.1016/j.nedt.2016.07.011
26Prati G., Albanesi C., Pietrantoni L.The Reciprocal Relationship between Sense of Community and Social Well-Being: A Cross-Lagged Panel Analysis2016Social Indicators Research10.1007/s11205-015-1012-8
27Roemer E., Henseler J.The dynamics of electric vehicle acceptance in corporate fleets: Evidence from Germany2022Technology in Society10.1016/j.techsoc.2022.101938
28Chaparro-Peláez J., Pereira-Rama A., Pascual-Miguel F.J.Inter-organizational information systems adoption for service innovation in building sector2014Journal of Business Research10.1016/j.jbusres.2013.11.026
29Lauro N.C., Grassia M.G., Cataldo R.Model-Based Composite Indicators: New Developments in Partial Least Squares-Path Modeling for the Building of Different Types of Composite Indicators2018Social Indicators Research10.1007/s11205-016-1516-x
30Zhu X., Wei Y., Lai Y., Li Y., Zhong S., Dai C.Empirical analysis of the driving factors of China’s ’Land finance’ mechanism using soft budget constraint theory and the PLS-SEM model2019Sustainability (Switzerland)10.3390/su11030742
31Lee W.-K.An elaboration likelihood model-based longitudinal analysis of attitude change during the process of IT acceptance via an education program2012Behaviour and Information Technology10.1080/0144929X.2010.547219
32Hallencreutz J., Parmler J.Important drivers for customer satisfaction–from a product focus to image and service quality2021Total Quality Management and Business Excellence10.1080/14783363.2019.1594756
33Guo Z., Tan F.B., Turner T., Xu H.Group norms, media preferences, and group meeting success: A longitudinal study2010Computers in Human Behavior10.1016/j.chb.2010.01.001
34Robina-Ramírez R., Medina Merodio J.A., McCallum S.What role do emotions play in transforming students’ environmental behavior at school?2020Journal of Cleaner Production10.1016/j.jclepro.2020.120638

References

  1. Sardana, K.; Gupta, T.; Kumar, B.; Gautam, H.K.; Garg, V.K. A cross-sectional pilot study of antibiotic resistance in Propionibacterium acnes strains in Indian acne patients using 16s-RNA polymerase chain reaction: A comparison among treatment modalities including antibiotics, benzoyl peroxide, and isotretinoin. Indian J. Dermatol. 2016, 61, 45–52. [Google Scholar] [CrossRef] [PubMed]
  2. Coughlin, S.S.; Datta, B.; Berman, A.; Hatzigeorgiou, C. A cross-sectional study of financial distress in persons with multimorbidity. Prev. Med. Rep. 2021, 23, 101464. [Google Scholar] [CrossRef] [PubMed]
  3. Shinde, S.; Setia, M.S.; Singh Setia, M.; Row-Kavi, A.; Anand, V.; Jerajani, H. Male sex workers: Are we ignoring a risk group in Male sex workers: Are we ignoring a risk group in Mumbai, India? Indian J. Dermatol. Venereol. Leprol. 2009, 75, 41–46. [Google Scholar]
  4. Cheung, M.L.; Pires, G.D.; Iii, P.J.R.; De Oliveira, M.J. Driving COBRAs: The power of social media marketing. Mark. Intell. Plan. 2020, 39, 361–376. [Google Scholar] [CrossRef]
  5. Milicev, J.; Qualter, P.; Goodfellow, C.; Inchley, J.; Simpson, S.; Leyland, A.; Karicha, K.; Long, E. The prospective relationship between loneliness, life satisfaction and psychological distress before and during the COVID-19 pandemic in the UK. J. Public Health 2022, 30, 2717. [Google Scholar] [CrossRef]
  6. Michel, J.S.; Rotch, M.A.; Carson, J.E.; Bowling, N.A.; Shifrin, N.V. Flattening the Latent Growth Curve? Explaining Within-Person Changes in Employee Well-Being during the COVID-19 Pandemic. Occup. Health Sci. 2021, 5, 247–275. [Google Scholar] [CrossRef] [PubMed]
  7. Szabó, A.; Sheridan, J.; Newcombe, D. Ten-Year Trajectories of Alcohol Consumption in Older Adult New Zealanders. J. Gerontol. Ser. B 2019, 76, 496–506. [Google Scholar] [CrossRef]
  8. Jeon, H.G.; Jeong, E.J.; Lee, S.J.; Kim, J.A. Exploring the Mechanism of Pathological Gaming in Adolescents: Focused on the Mediation Paths and Latent Group Comparison. Front. Psychol. 2022, 12, 756328. [Google Scholar] [CrossRef]
  9. Batchelder, A.W.; Glynn, T.R.; Moskowitz, J.T.; Neilands, T.B.; Dilworth, S.; Rodriguez, S.L.; Carrico, A.W. The shame spiral of addiction: Negative self-conscious emotion and substance use. PLoS ONE 2022, 17, e0265480. [Google Scholar] [CrossRef]
  10. Ho, T.T.H.; Le, V.H.; Nguyen, D.T.; Nguyen, C.T.P.; Nguyen, H.T.T. Effects of career development learning on students’ perceived employability: A longitudinal study. High. Educ. 2022, 1–19. [Google Scholar] [CrossRef]
  11. Wu, W.; West, S.G.; Taylor, A.B. Evaluating model fit for growth curve models: Integration of fit indices from SEM and MLM frameworks. Psychol. Methods 2009, 14, 183–201. [Google Scholar] [CrossRef] [PubMed]
  12. Lee, T.K.; Wickrama, K.K.A.S.; O’Neal, C.W. Application of Latent Growth Curve Analysis With Categorical Responses in Social Behavioral Research. Struct. Equ. Model. A Multidiscip. J. 2017, 25, 294–306. [Google Scholar] [CrossRef] [PubMed]
  13. Isiordia, M.; Ferrer, E. Curve of Factors Model: A Latent Growth Modeling Approach for Educational Research. Educ. Psychol. Meas. 2016, 78, 203–231. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  14. Zhang, D.; Huebner, E.S.; Tian, L. Neuroticism and cyberbullying among elementary school students: A latent growth curve modeling approach. Pers. Individ. Differ. 2020, 171, 110472. [Google Scholar] [CrossRef]
  15. Tomasik, M.J.; Helbling, L.A.; Moser, U. Educational gains of in-person vs. distance learning in primary and secondary schools: A natural experiment during the COVID-19 pandemic school closures in Switzerland. Int. J. Psychol. 2020, 56, 566–576. [Google Scholar] [CrossRef] [PubMed]
  16. Birkeland, M.S.; Holt, T.; Ormhaug, S.M.; Jensen, T.K. Perceived social support and posttraumatic stress symptoms in children and youth in therapy: A parallel process latent growth curve model. Behav. Res. Ther. 2020, 132, 103655. [Google Scholar] [CrossRef]
  17. Curran, P.J.; Obeidat, K.; Losardo, D. Twelve Frequently Asked Questions About Growth Curve Modeling. J. Cogn. Dev. 2010, 11, 121–136. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  18. Felt, J.; Depaoli, S.; Tiemensma, J. Latent Growth Curve Models for Biomarkers of the Stress Response. Front. Neurosci. 2017, 11, 315. [Google Scholar] [CrossRef] [Green Version]
  19. Fornell, C.; Bookstein, F.L. Two Structural Equation Models: LISREL and PLS Applied to Consumer Exit-Voice Theory. J. Mark. Res. 1982, 19, 440. [Google Scholar] [CrossRef] [Green Version]
  20. Willaby, H.W.; Costa, D.S.; Burns, B.D.; MacCann, C.; Roberts, R.D. Testing complex models with small sample sizes: A historical overview and empirical demonstration of what Partial Least Squares (PLS) can offer differential psychology. Pers. Individ. Differ. 2015, 84, 73–78. [Google Scholar] [CrossRef]
  21. Hair, J.F., Jr.; Matthews, L.M.; Matthews, R.L.; Sarstedt, M. PLS-SEM or CB-SEM: Updated guidelines on which method to use. Int. J. Multivar. Data Anal. 2017, 1, 107. [Google Scholar] [CrossRef]
  22. Roemer, E. A tutorial on the use of PLS path modeling in longitudinal studies. Ind. Manag. Data Syst. 2016, 116, 1901–1921. [Google Scholar] [CrossRef]
  23. Phulwani, P.R.; Kumar, D.; Goyal, P. A Systematic Literature Review and Bibliometric Analysis of Recycling Behavior. J. Glob. Mark. 2020, 33, 354–376. [Google Scholar] [CrossRef]
  24. Nagariya, R.; Kumar, D.; Kumar, I. Service supply chain: From bibliometric analysis to content analysis, current research trends and future research directions. Benchmarking 2020, 28, 333–369. [Google Scholar] [CrossRef]
  25. Linnenluecke, M.K.; Marrone, M.; Singh, A.K. Conducting systematic literature reviews and bibliometric analyses. Aust. J. Manag. 2019, 45, 175–194. [Google Scholar] [CrossRef]
  26. Muñoz, E.; Robins, R.W.; Sutin, A.R. Perceived ethnic discrimination and cognitive function: A 12-year longitudinal study of Mexican-origin adults. Soc. Sci. Med. 2022, 311, 115296. [Google Scholar] [CrossRef]
  27. Yarrington, J.S.; Vinograd, M.; Williams, A.L.; Wolitzky-Taylor, K.B.; Zinbarg, R.E.; Mineka, S.; Waters, A.M.; Craske, M.G. Fear-potentiated startle predicts longitudinal change in transdiagnostic symptom dimensions of anxiety and depression. J. Affect. Disord. 2022, 311, 399–406. [Google Scholar] [CrossRef]
  28. Russo, S.; Colloca, P.; Cavazza, N.; Roccato, M. Household crowding during the COVID-19 lockdown fosters anti-democracy even after 17 months: A 5-wave latent growth curve study. J. Environ. Psychol. 2022, 83, 101867. [Google Scholar] [CrossRef]
  29. McArdle, J.J. Dynamic but Structural Equation Modeling of Repeated Measures Data. In Handbook of Multivariate Experimental Psychology; Nesselroade, J.R., Cattell, R.B., Eds.; Springer: New York, USA, 1988; pp. 561–614. [Google Scholar]
  30. McArdle, J.J. Latent variable growth within behavior genetic models. Behav. Genet. 1986, 16, 163–200. [Google Scholar]
  31. Bollen, K.A.; Curran, P.J. Latent Curve Models: A Structural Equation Perspective; Wiley-Interscience: Hoboken, NJ, USA, 2006. [Google Scholar]
  32. Wold, H. Path Models with Latent Variables: The NIPALS Approach. In Quantitative Sociology; Academic Press: Cambridge, MA, USA, 1975; pp. 307–357. [Google Scholar] [CrossRef]
  33. Wold, H. Partial Least Squares. In Encyclopedia of Statistical Sciences; Kotz, S., Johnson, N.L., Eds.; John Wiley: New York, NY, USA, 1985; pp. 581–591. [Google Scholar]
  34. Lohmoller, J.-B.; Berlin, S.-V.; Gmbh, H. Latent Variable Path Modeling with Partial Least Squares; Physica-Verlag Heidelberg: Heidelberg, Germany, 1989. [Google Scholar]
  35. Henseler, J.; Hubona, G.; Ray, P.A. Using PLS path modeling in new technology research: Updated guidelines. Ind. Manag. Data Syst. 2016, 116, 2–20. [Google Scholar] [CrossRef]
  36. Chin, W.W. The Partial Least Squares Approach to Structural Equation Modeling. Available online: https://www.researchgate.net/publication/311766005 (accessed on 9 November 2022).
  37. Nitzl, C.; Chin, W.W. The case of partial least squares (PLS) path modeling in managerial accounting research. J. Manag. Control. 2017, 28, 137–156. [Google Scholar] [CrossRef]
  38. Haenlein, M.; Kaplan, A.M. A Beginner’s Guide to Partial Least Squares Analysis. Underst. Stat. 2004, 3, 283–297. [Google Scholar] [CrossRef]
  39. Shea, C. Efficacy-performance spirals: An empirical test. J. Manag. 2000, 26, 791–812. [Google Scholar] [CrossRef]
  40. Hennig-Thurau, T.; Groth, M.; Paul, M.; Gremler, D.D. Are All Smiles Created Equal? How Emotional Contagion and Emotional Labor Affect Service Relationships. J. Mark. 2006, 70, 1547–7185. Available online: http://www.marketingpower.com/jmblog (accessed on 14 November 2022). [CrossRef]
  41. Johnson, M.D.; Herrmann, A.; Huber, F. The Evolution of Loyalty Intentions. J. Mark. 2006, 70, 122–132. [Google Scholar] [CrossRef] [Green Version]
  42. Jones, E.; Sundaram, S.; Chin, W. Factors Leading to Sales Force Automation Use: A Longitudinal Analysis. J. Pers. Sell. Sales Manag. 2002, 22, 145–156. [Google Scholar]
  43. Jacobs, N.; Hagger, M.S.; Streukens, S.; De Bourdeaudhuij, I.; Claes, N. Testing an integrated model of the theory of planned behaviour and self-determination theory for different energy balance-related behaviours and intervention intensities. Br. J. Health Psychol. 2011, 16, 113–134. [Google Scholar] [CrossRef]
  44. Inamdar, Z.; Raut, R.; Narwane, V.S.; Gardas, B.; Narkhede, B.; Sagnak, M. A systematic literature review with bibliometric analysis of big data analytics adoption from period 2014 to 2018. J. Enterp. Inf. Manag. 2020, 34, 101–139. [Google Scholar] [CrossRef]
  45. Okoli, C.; Schabram, K. A Guide to Conducting a Systematic Literature Review of Information Systems Research. Work. Pap. Inf. Systems 2010, 10, 2010. [Google Scholar] [CrossRef] [Green Version]
  46. Donthu, N.; Kumar, S.; Mukherjee, D.; Pandey, N.; Lim, W.M. How to conduct a bibliometric analysis: An overview and guidelines. J. Bus. Res. 2021, 133, 285–296. [Google Scholar] [CrossRef]
  47. Xiao, Y.; Watson, M. Guidance on Conducting a Systematic Literature Review. J. Plan. Educ. Res. 2017, 39, 93–112. [Google Scholar] [CrossRef]
  48. Lockwood, C.; Munn, Z.; Porritt, K. Qualitative research synthesis. Int. J. Evid. Based Healthc. 2015, 13, 179–187. [Google Scholar] [CrossRef] [PubMed]
  49. Shaffril, H.A.M.; Ahmad, N.; Samsuddin, S.F.; Abu Samah, A.; Hamdan, M.E. Systematic literature review on adaptation towards climate change impacts among indigenous people in the Asia Pacific regions. Int. J. Evid. Based Healthc. 2015, 39, 879–910. [Google Scholar] [CrossRef]
  50. Zhu, J.; Liu, W. A tale of two databases: The use of Web of Science and Scopus in academic papers. Scientometrics 2020, 123, 321–335. [Google Scholar] [CrossRef] [Green Version]
  51. Liu, W. The data source of this study is Web of Science Core Collection? Not enough. Scientometrics 2019, 121, 1815–1824. [Google Scholar] [CrossRef]
  52. Li, K.; Rollins, J.; Yan, E. Web of Science use in published research and review papers 1997–2017: A selective, dynamic, cross-domain, content-based analysis. Scientometrics 2017, 115, 1–20. [Google Scholar] [CrossRef] [Green Version]
  53. Harzing, A.-W.; Alakangas, S. Google Scholar, Scopus and the Web of Science: A longitudinal and cross-disciplinary comparison. Scientometrics 2016, 106, 787–804. [Google Scholar] [CrossRef]
  54. Feng, Y.; Zhu, Q.; Lai, K.-H. Corporate social responsibility for supply chain management: A literature review and bibliometric analysis. J. Clean. Prod. 2017, 158, 296–307. [Google Scholar] [CrossRef]
  55. Pranckutė, R. Web of Science (WoS) and Scopus: The Titans of Bibliographic Information in Today’s Academic World. Publications 2021, 9, 12. [Google Scholar] [CrossRef]
  56. Okoli, C. A Guide to Conducting a Standalone Systematic Literature Review. Commun. Assoc. Inf. Syst. 2015, 37, 879–910. [Google Scholar] [CrossRef] [Green Version]
  57. Levy, Y.; Ellis, T.J. A Systems Approach to Conduct an Effective Literature Review in Support of Information Systems Research. Informing Sci. Int. J. Emerg. Transdiscipl. 2006, 9, 181–212. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  58. Brereton, P.; Kitchenham, B.A.; Budgen, D.; Turner, M.; Khalil, M. Lessons from applying the systematic literature review process within the software engineering domain. J. Syst. Softw. 2006, 80, 571–583. [Google Scholar] [CrossRef]
  59. Pritchard, A. Statistical Bibliography or Bibliometrics? J. Doc. 1969, 25, 348–349. [Google Scholar]
  60. Vogel, R.; Güttel, W.H. The Dynamic Capability View in Strategic Management: A Bibliometric Review. Int. J. Manag. Rev. 2013, 15, 426–446. [Google Scholar] [CrossRef]
  61. Liang, T.-P.; Liu, Y.-H. Research Landscape of Business Intelligence and Big Data analytics: A bibliometrics study. Expert Syst. Appl. 2018, 111, 2–10. [Google Scholar] [CrossRef]
  62. Cobo, M.J.; López-Herrera, A.G.; Herrera-Viedma, E.; Herrera, F. An approach for detecting, quantifying, and visualizing the evolution of a research field: A practical application to the Fuzzy Sets Theory field. J. Informetr. 2011, 5, 146–166. [Google Scholar] [CrossRef]
  63. Xu, X.; Chen, X.; Jia, F.; Brown, S.; Gong, Y.; Xu, Y. Supply chain finance: A systematic literature review and bibliometric analysis. Int. J. Prod. Econ. 2018, 204, 160–173. [Google Scholar] [CrossRef]
  64. Sarstedt, M.; Ringle, C.M.; Hair, J.F. Handbook of Market Research; Springer: Cham, Switzerland, 2020. [Google Scholar] [CrossRef]
  65. Limayem, M.; Cheung, C.M. Understanding information systems continuance: The case of Internet-based learning technologies. Inf. Manag. 2008, 45, 227–232. [Google Scholar] [CrossRef]
  66. Islam, A.N. Investigating e-learning system usage outcomes in the university context. Comput. Educ. 2013, 69, 387–399. [Google Scholar] [CrossRef]
  67. Chen, C.-P.; Lai, H.-M.; Ho, C.-Y. Why do teachers continue to use teaching blogs? The roles of perceived voluntariness and habit. Comput. Educ. 2015, 82, 236–249. [Google Scholar] [CrossRef]
  68. Guo, Z.; Tan, F.B.; Turner, T.; Xu, H. Group norms, media preferences, and group meeting success: A longitudinal study. Comput. Hum. Behav. 2010, 26, 645–655. [Google Scholar] [CrossRef]
  69. Roxas, B. Effects of entrepreneurial knowledge on entrepreneurial intentions: A longitudinal study of selected South-east Asian business students. J. Educ. Work. 2013, 27, 432–453. [Google Scholar] [CrossRef]
  70. Roemer, E.; Henseler, J. The dynamics of electric vehicle acceptance in corporate fleets: Evidence from Germany. Technol. Soc. 2022, 68, 101938. [Google Scholar] [CrossRef]
  71. Robina-Ramírez, R.; Merodio, J.A.M.; McCallum, S. What role do emotions play in transforming students’ environmental behaviour at school? J. Clean. Prod. 2020, 258, 120638. [Google Scholar] [CrossRef]
  72. Keil, M.; Tan, B.C.Y.; Wei, K.-K.; Saarinen, T.; Tuunainen, V.; Wassenaar, A. A Cross-Cultural Study on Escalation of Commitment Behavior in Software Projects. Manag. Inf. Syst. Q. 2000, 24, 299–323. [Google Scholar] [CrossRef]
  73. Williams, L.J.; Vandenberg, R.J.; Edwards, J.R. 12 Structural Equation Modeling in Management Research: A Guide for Improved Analysis. Acad. Manag. Ann. 2009, 3, 543–604. [Google Scholar] [CrossRef]
  74. Crowley, S.L.; Fan, X. Structural Equation Modeling: Basic Concepts and Applications in Personality Assessment Research. J. Pers. Assess. 1997, 68, 508–531. [Google Scholar] [CrossRef]
  75. Wright, S. The method of path coefficients. Ann. Math. Stat. 1934, 5, 161–215. [Google Scholar] [CrossRef]
  76. Spearman, C. “General Intelligence”, Objectively Determined and Measured. Am. J. Psychol. 1904, 15, 201–292. [Google Scholar] [CrossRef]
  77. Preacher, K.J.; Wichman, A.L.; Maccallum, R.C.; Briggs, N.E. Latent Growth Curve Modeling; SAGE Publications, Inc.: Thousand Oaks, CA, USA, 2008. [Google Scholar]
  78. Mittal, V.; Kumar, P.; Tsiros, M. Attribute-Level Performance, Satisfaction, and Behavioral Intentions over Time: A Consumption-System Approach. J. Mark. 1999, 63, 88–101. [Google Scholar] [CrossRef]
Data 08 00032 g001 550
Figure 1. The Flow Diagram of Reviewing Process.
Figure 1. The Flow Diagram of Reviewing Process.
Data 08 00032 g001
Data 08 00032 g002 550
Figure 2. Annual growth of publications (Scopus & WoSCC databases, 1986–2022).
Figure 2. Annual growth of publications (Scopus & WoSCC databases, 1986–2022).
Data 08 00032 g002
Data 08 00032 g003 550
Figure 3. Co-authorship network map in the field of panel survey data.
Figure 3. Co-authorship network map in the field of panel survey data.
Data 08 00032 g003
Data 08 00032 g004 550
Figure 4. Network visualization map of citations based on the documents.
Figure 4. Network visualization map of citations based on the documents.
Data 08 00032 g004
Data 08 00032 g005 550
Figure 5. Network visualization map of co-citation of cited authors.
Figure 5. Network visualization map of co-citation of cited authors.
Data 08 00032 g005
Data 08 00032 g006 550
Figure 6. The network visualization map of co-occurrences of keywords.
Figure 6. The network visualization map of co-occurrences of keywords.
Data 08 00032 g006
Data 08 00032 g007 550
Figure 7. The overlay visualization map of keywords according to year.
Figure 7. The overlay visualization map of keywords according to year.
Data 08 00032 g007
Table
Table 1. Search string for the retrieved records.
Table 1. Search string for the retrieved records.
DatabaseSearch String
ScopusTITLE-ABS-KEY(("panel survey" OR "longitudinal survey" OR "panel data" OR "longitudinal") AND ("partial least squares" OR "latent growth curve" OR "LGCM" OR "PLS Path" OR "PLS-SEM"))
WoSCCTS=(("panel survey" OR "longitudinal survey" OR "panel data" OR "longitudinal") AND ("partial least squares" OR "latent growth curve" OR "LGCM" OR "PLS Path" OR "PLS-SEM"))
Table
Table 2. Inclusion and Exclusion Criteria.
Table 2. Inclusion and Exclusion Criteria.
Database CriteriaInclusionExclusion
TimelineAll records in Scopus and WoSCC databases.Other databases.
LanguageEnglish.Other languages.
Document TypeArticle, Article review, and Conference.Books and chapters in a book.
Subject areaPsychology, Social Sciences, Business, Management, Accounting, Mathematics, Economics, and Multidisciplinary, Behavioral SciencesOther subject areas in bibliographic databases of Scopus and WoSCC.
MethodLGCM, PLS-SEM, and Partial Least Squares.Multilevel Linear Growth Curve Model, Bayesian Growth Curve Model, Repeated Measure ANOVA, Generalized estimating equations, and Mixed effect regression.
Type of dataLongitudinal survey and panel survey data.Cross-sectional data.
Table
Table 3. Top 10 journals contributing to the panel survey data.
Table 3. Top 10 journals contributing to the panel survey data.
Source (Journal)Total
Publications		Total
Citations
Developmental Psychology514231
Structural Equation Modeling301267
Journal of Youth and Adolescence29930
PLoS ONE28498
Psychology and Aging251061
Journal of Affective Disorders25234
Journal of Abnormal Child Psychology221043
Journals of Gerontology21584
Frontiers in Psychology20200
Journal of Adolescence19741
Table
Table 4. The most cited articles related to LGCM.
Table 4. The most cited articles related to LGCM.
RankAuthorsYearDOICitations
1McArdle J.J., Epstein D.198710.2307/1130295653
2Ge X., Lorenz F.O., Conger R.D., Elder Jr. G.H., Simons R.L.199410.1037/0012-1649.30.4.467625
3Plutzer E.200210.1017/S0003055402004227549
4McArdle J.J., Ferrer-Caja E., Hamagami F., Woodcock R.W.200210.1037/0012-1649.38.1.115401
5Wang M.200710.1037/0021-9010.92.2.455388
Table
Table 5. The most cited article related to PLS-SEM.
Table 5. The most cited article related to PLS-SEM.
RankAuthorsYearDOICitations
1Limayem M., Cheung C.M.K.200810.1016/j.im.2008.02.005369
2Baer J.S., Sampson P.D., Barr H.M., Connor P.D., Streissguth A.P.200310.1001/archpsyc.60.4.377284
3Wong V.W.-S., Tse C.-H., Lam T.T.-Y., Wong G.L.-H.201310.1371/jounal.pone.0062885217
4Dodge K.A., Malone P.S., Lansford J.E., Shari M., Pettit G.S., Bates.200910.1111/j.15405834.2009.00528.x210
5Hennig-Thurau T., Henning V., Sattler H.200710.1509/jmkg.71.4.001208
Table
Table 6. Co-occurrence of author keyword in panel survey data.
Table 6. Co-occurrence of author keyword in panel survey data.
Cluster 1 (Red)Cluster 2 (Green)Cluster 3 (Blue)
Mental Health (27)Developmental Trajectories (48)Longitudinal Study (292)
Self-Efficacy (18)Gender (47)Aging (22)
Social Support (18)Personality Development (38)Older Adults (15)
Cognitive Aging (16)Parenting (26)Cognition (14)
PLS-SEM (14)Substance Use (20)Psychological Well-Being (12)
Adoption (12)Academic Achievement (15)Dementia (11)
Life Satisfaction (12)Growth Curve Modeling (14)Partial Least Squares (11)
Stress (12)Motivation (14)Cluster 6 (Light Blue)
Bullying (11)Effortful Control (10)Emerging Adulthood (16)
Education (11)Self-Regulation (10)Delinquency (12)
Job Satisfaction (10)Well-Being (10)Cluster 7 (Orange)
Cluster 4 (Yellow)Cluster 5 (Purple)Latent Growth Curve Model (294)
Depression (91)Adolescence (169)Children (10)
Trajectories (32)Alcohol (32)Cluster 8 (Brown)
Depressive Symptom (31)Physical Activity (19)Satisfaction (10)
Anxiety (28)Aggression (10)Social Media (10)
Self-Esteem (18)Smoking (10)Cluster 9 (Pink)
Life Course (12) Structural Equation Modeling (22)
Loneliness (12) COVID-19 (15)
Table
Table 7. Summary of approaches practiced in analyzing a structural model.
Table 7. Summary of approaches practiced in analyzing a structural model.
Type of ModelDescriptionsLimitationsAuthors
Model 1:
Pre and Post approach with different construct.
  • Two periods of time.
  • Pre and Post approach with different constructs.
  • Used path analysis.
  • Not suitable for more than two periods of time.
  • Cannot measure the evaluation of effect over time.
[65,66,67]
Model 2:
Path Comparison approach.
  • Two periods of time.
  • Using the same construct at the first and the second time of survey.
  • Analyze two models separately according to time (t0 and t1).
  • Used path analysis.
  • Comparing these two models using a t-test.
  • Not suitable for more than two periods of time.
  • Cannot evaluate the changes in one structural model.
  • ∙
[68]
Model 3:
Cross-lagged approach.
  • Two periods of time.
  • Using the same construct at the first and the second time of survey.
  • Used Cross-lagged approach.
  • Not suitable for more than two periods of time.
  • Cannot assess the growth trajectories.
  • Required a few assumptions.
[69]
Model 4:
Pre and Post approach with same construct.
  • Two periods of time.
  • Used paired t-test for evaluating the differences between indicators (t1 and t2).
  • Evaluating the effect between changes of constructs based on the value of differences between indicators.
  • Used path analysis.
  • Not suitable for more than two periods of time.
  • Cannot assess the growth of trajectories.
[43]
Model 5:
Evaluation approach.
  • More than two periods of time.
  • Measured direct effect and carry-over effect.
  • Used paired t-test, path analysis, and bias-corrected confidence interval.
  • Do not have one structural model.
  • Cannot assess model fit.
  • Cannot assess the whole changes in one structural model.
  • Cannot assess individual trajectories and factors influencing those changes simultaneously.
[41,70,71]
Table
Table 8. The procedure of the PLS-SEM approach in analyzing panel survey data.
Table 8. The procedure of the PLS-SEM approach in analyzing panel survey data.
Type of ModelProcedureArticles
Model 1:
Pre and Post approach with different constructs.		Data 08 00032 i001[65,66,67]
Model 2:
Path Comparison approach.		Data 08 00032 i002[68]
Model 3:
Cross-lagged approach.		Data 08 00032 i003[69]
Model 4:
Pre and Post approach with same construct.		Data 08 00032 i004[43]
Model 5:
Evaluation Approach.		Data 08 00032 i005[41,70,71]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Mohd Ghazali, Z.; Wan Yaacob, W.F.; Wan Omar, W.M. LGCM and PLS-SEM in Panel Survey Data: A Systematic Review and Bibliometric Analysis. Data 2023, 8, 32. https://doi.org/10.3390/data8020032

AMA Style

Mohd Ghazali Z, Wan Yaacob WF, Wan Omar WM. LGCM and PLS-SEM in Panel Survey Data: A Systematic Review and Bibliometric Analysis. Data. 2023; 8(2):32. https://doi.org/10.3390/data8020032

Chicago/Turabian Style

Mohd Ghazali, Zulkifli, Wan Fairos Wan Yaacob, and Wan Marhaini Wan Omar. 2023. "LGCM and PLS-SEM in Panel Survey Data: A Systematic Review and Bibliometric Analysis" Data 8, no. 2: 32. https://doi.org/10.3390/data8020032

APA Style

Mohd Ghazali, Z., Wan Yaacob, W. F., & Wan Omar, W. M. (2023). LGCM and PLS-SEM in Panel Survey Data: A Systematic Review and Bibliometric Analysis. Data, 8(2), 32. https://doi.org/10.3390/data8020032

Article Metrics

Back to TopTop