Visual versus Tabular Scheduling Programs

Abstract

Effective scheduling in construction is crucial for ensuring timely project completion and maintaining budget control. Scheduling programs play an important role in this process by providing digital tools to develop, monitor, and adjust project timelines effectively. In the industry’s current method of practice, tabular scheduling programs are utilized that require users to enter task information and their relationships in a tabular format. Recently, a new scheduling program approach called visual scheduling has emerged that requires users to draw the network diagram to create the schedule. This paper presents an experimental study that evaluated the efficiency of schedule creation using a visual scheduling program compared to two tabular scheduling programs. The results show that the time spent creating a schedule using the visual scheduling program was significantly shorter than using tabular scheduling programs. Participants found visual scheduling easier to define tasks, define correlations, spot mistakes in the schedule, make changes to the schedule, and understand the overall schedule. The majority of the participants reported visual scheduling as a tool that allowed them to create schedules faster. They also found it to be a more intuitive scheduling tool and a method that can reduce the possibility of making mistakes during scheduling.

1. Introduction

Successful project execution relies heavily on strong project management, with scheduling serving as a key element. Scheduling is essential to ensure projects are executed as planned making it a foundational aspect of project management and analysis [1]. By offering a detailed timeline for project deliverables, scheduling helps in structuring tasks to keep them organized and manageable [2]. Without effective scheduling, project timelines can become unpredictable, resulting in hurried tasks and comprised quality [3]. In the world of project management, computer applications play an increasing role in project deliveries. To support the complexities of organizing, planning, and managing resources, Project Management Software (PMS) packages are indispensable. These tools not only handle scheduling and budgeting but also facilitate resource assignment resulting in cost-optimal scheduling [4] and surpassing manual methods in performance [5]. The adoption of PMS has become progressively prevalent across various industries due to its significant advantages. An increasing number of companies acknowledge the advantages of systematic project management, such as reduced costs, enhanced efficiency, improved satisfaction among customers and stakeholders, and a higher competitive advantage in the industry [6].

Construction projects are known for their substantial complexity and high levels of risk, often leading to extended timelines and multiple phases. The inherent unpredictability of these projects often causes errors, risks, and modifications due to the involvement of various stakeholders and the nature of construction work plans. Consequently, the use of PMS is crucial in the construction industry to manage these challenges effectively [7]. The most commonly reported factors influencing the use of PMS are the size and complexity of projects, which compel project management professionals to adopt these tools more extensively [8]. Consequently, some small and medium-scale construction companies confront various issues such as inefficient planning, project delays, and inefficiency of resources due to reluctance to adopt digital tools [9]. As huge users of PMS, construction industry professionals are interested in improving the tools and techniques available for better project planning and control [10]. However, not all users are concerned with the most sophisticated package with advanced tools and features. They mostly look for programs that allow them to produce simple time charts, simple resource and cost analyses, and basic reports [11]. Studies have demonstrated success in implementing PMS in construction projects for scheduling in residential [12,13] and commercial projects such as hotels, schools, and high-rise buildings [14,15,16]. These packages have shown success in resource allocation and leveling in schedules [17]. Selecting the appropriate PMS to use is crucial as it significantly impacts project efficiency and control [18], which has motivated researchers to evaluate them based on various project management criteria [19]. A survey conducted with eighty construction professionals in 2020 revealed relatively low user satisfaction with available PMS packages as they do not integrate all the construction aspects, making it difficult for construction companies to select the most appropriate software. Some participants indicated using more than one package to address their project needs [20].

In the realm of project management, visual explanations and tools can demonstrate their profound impact on enhancing understanding and efficiency. Translating verbal information from the text into a visual explanation can promote deeper processing of the material and more complete and comprehensive mental models [21]. Visual representations ensure all components of a system are included and work together properly, promoting overall coherence and effectiveness in achieving project goals [22]. Visual approaches not only aid in mastering complex systems but also facilitate better project scheduling and management practices, as the clarity and depth of understanding are significantly enhanced through visual means.

In contrast to purely symbolic words, visual communications can convey some content and structure directly [23]. A sketch can enable designers to visualize new configurations and generate ideas that may not have been anticipated during the initial creation [24]. This capability is crucial in project management, where diagrams and sketches facilitate inferences by leveraging physical features, such as proximity, angle, and connectivity, and capture complex relations between components [25]. Problem-solving can proceed through a smooth transit of the diagrams [26]. Therefore, as the construction industry increasingly adopts PMS, the integration of visual tools and methodologies becomes essential, not only for improving project efficiency and control but also for enhancing learning outcomes and problem-solving capabilities in project scheduling.

One of the methods for visualizing schedules is the Gantt chart, which originated with Gantt’s work during the First World War [27]. It was introduced as a production planning tool but not adopted for widespread use in construction management until the 1980s [28]. Gantt charts are simple bar charts displayed on a time scale, where each bar represents a task with its start and end times indicated by the bar’s edges [29]. It is the most common visual method for scheduling because it is easy to prepare, read, and understand [30]. Consequently, early PMS packages allowed users to present their results in the form of Gantt charts [31,32], which are still used in current packages. Despite their readability and simplicity, they have various disadvantages for large projects, including the lack of precedence relationships between the tasks [33]. Thus, they are not ideal as primary tools for project monitoring and management [28].

The Critical Path Method (CPM) was developed by Du Pont and UNTVAC in 1957. Since then, it has become a very popular tool for scheduling complex and large projects, which is the basis for the analysis of most PMS packages [34]. This method is based on a network diagram, which is the most popular visual platform for presenting project schedules [28,35]. Network diagrams feature nodes, arrows, and labels to represent task durations and logical constraints. The activity-on-node (AON) method is the most common form of network diagram representation, where nodes denote tasks and arrows illustrate the logical relationships between them [28]. Analyzing this diagram identifies the critical path, which is the longest path from the start to end milestones and indicates the minimum time required to complete the project [34]. Although the Gantt chart lacks the ability to represent logical dependencies compared to the network diagram, it is one of the most powerful tools used in the industry. Based on a series of interviews conducted with project managers in 2024, the Gantt chart was reported to have a crucial role in project management by offering a clear visualization of the project schedule, enabling team members to easily understand timelines, making it indispensable for effective project management [36]. According to a survey of seven hundred fifty project managers in 2008, the Gantt chart was the fourth most frequently used tool out of seventy different project management tools and techniques, while the network diagram was ranked forty-second [37].

The majority of the current commercially available PMS packages, such as Microsoft Project [38], Oracle Primavera P6 [39], and Procore [40], use the tabular scheduling method. Users enter data such as task and constraint information in a tabular format, and the program generates a Gantt chart next to the table by default settings. With advances in technology, most recently, PMS packages have been developed based on the visual scheduling method, which allows users to draw the network diagram to create the schedule. In the current method of practice in the industry, before entering any data into the programs, contractors draw network diagrams to brainstorm project schedules by using sticky notes on the board that show the tasks and their duration and draw lines between them to show the relationships [41,42]. This brings up the question of whether the visual scheduling method in PMS packages could be more efficient for creating schedules as it mirrors the way professionals perform scheduling in practice. This study aimed to examine the efficiency of the visual scheduling methods for schedule creation without the consideration of cost or resource constraints.

2. Materials and Methods

To evaluate the efficiency of the visual scheduling compared to the tabular scheduling method, an experiment was conducted with one visual and two tabular scheduling programs. The two tabular scheduling programs in this study were among the most frequently used programs in the construction industry and were set as the comparison baseline. A start-up company provided the visual scheduling program. In this paper, all three program names are kept anonymous; the visual scheduling program is called “Vis”, and the two tabular scheduling programs are called “Tab1” and “Tab2”. The experiment required participants to create schedules for three comparable project scenarios with all three programs assigned randomly. The time participants took to complete each scenario was measured for quantitative analysis. The user feedback on their experience with each program and their comparison was collected with questionnaires. The following sections provide detailed information about the participants, experimental setup, scenarios, questionnaires, and experimental procedure.

2.1. Participants

The study involved 55 graduate students enrolled in the CMGT 650 construction project management, planning, and scheduling course in the construction management master’s program at California State University, East Bay. The average age was 25 years, with 30.9% female and 69.1% male. The majority of the participants, 92.7%, had a bachelor’s degree in architecture, engineering, or construction (AEC). They had an average of two years of AEC industry experience. Among the participants, 21.8% had prior experience using Tab1, 16.3% with Tab2, and 12.7% had used visual scheduling programs similar to Vis. Throughout the semester, participants were trained on using Tab1, Tab2, and Vis in lab sessions for 9 weeks and completed assignments with these programs. They were also required to submit a team project using all three programs. This ensured that participants were well-prepared to work with these scheduling programs by the time of the experiment, which was the end of the semester.

2.2. Experiment Setup

The experiment took place in a computer laboratory with Tab1 and Tab2 installed. Participants were provided access to Vis through individual user accounts. Computers had access to the internet and allowed participants to log into Vis, a cloud-based program. Programs Tab1 and Tab2 have relatively similar tabular interfaces. The task names are entered into one column, and the information correlated to the tasks can be seen on the row in front of each task name cell. The duration and correlation of each task with others are entered in a tabular interface. With the default settings, these programs generate a Gantt chart next to the table. To view the network diagram, users need to change the settings to see it in another interface. On the other hand, the visual scheduling program Vis requires users to draw the network diagram on an empty digital page by utilizing drawing tools to create the schedule. Users define tasks by drawing a shape and typing the task name within the shape. They define the correlation between the tasks by drawing an arrow between the task shapes and entering the correlation information. To view the Gantt chart, users need to change the settings to see it in another interface that includes a table of task information similar to tabular scheduling programs. Participants were asked to use all the programs in an auto-schedule setting, meaning the programs automatically created the schedule as the information was entered.

2.3. Scenarios

Three scheduling scenarios, Sc1, Sc2, and Sc3, were designed to mirror one another in complexity. Each scenario consisted of the same number of activities, each activity with the same duration and with similar network relationships, lags, and relationship types. However, the scenarios differed in the sequencing of activities, resulting in different end dates. All scenarios had start and end milestones, along with 10 activities named with letters. This way, the time spent on typing the activity names was the same for all scenarios. Each activity had the same duration of 5 days, milestones had a duration of zero, and the lag duration was the same for all scenarios. As a result, the impact of entering a wrong activity or lag duration on the end date was minimized, meaning participants mainly had to check their relationship definitions in the programs to find possible scheduling mistakes. There were four different types of activity relationships: finish to start (FS), fish to start with a lag (FS + L), start to start with lag (SS + L), and finish to finish (FF). All scenarios had 15 relationships, where 9 of them were FS, 2 of them were SS + L, 3 of them were FS + L, and one of them was FF. All scenarios had a start date of 1 February 2024. To assist in validating the accuracy of the schedule, the end date was given to the participants. Participants were given the description of the scenarios on paper with no network diagram. They had the option to draw the network diagram next to the description; however, it was not mandatory. The descriptions of the scenarios are as follows.

2.3.1. Scenario 1

After the Notice to Proceed (start milestone), activities A and B can start (FS). Activity C will start one day after activity B starts (SS + 1). Activity D starts after activity C is finished (FS). Activity E will start two days after activity B is finished (FS + 2). Activity F will start two days after activity D is finished (FS + 2). Activities G and I will start once activity A is finished (FS). Additionally, activities G and I must finish at the same time (FF). Activity H starts one day after activity I starts (SS + 1). Activity J will start two days after activity H is finished (FS + 2) and after activity G is finished (FS). The Project Closeout (end milestone) (project closeout) will be completed once activities F, E, and J are finished (FS). After the completion of this schedule with the assignment program, the end date of February 26, 2024 (26 February 2024) will be reached. Figure 1 shows the network diagram created based on the description of the first scenario, Sc1.

2.3.2. Scenario 2

After the Notice to Proceed (start milestone), activity A can start (FS). Activities B and C will start once activity A is finished (FS). Additionally, activities B and C must finish at the same time (FF). Activity D will start two days after activity C finishes (FS + 2). Activity E will start one day after activity D starts (SS + 1). Activity F will start two days after activity E finishes (FS + 2). Activity G will start after activity B is done (FS). Activity H will start two days after activity G is over (FS + 2). Activity I will start one day after activity G starts (SS + 1). Activity J will start after activity I and H are finished (FS). The Project Closeout (end milestone) will be completed once activities J, D, and F are finished (FS). After the completion of the schedule with the assigned program, the end date of 8 March 2024 (03/08/2024) will be reached. Figure 2 shows the network diagram created based on the description of the second scenario, Sc2.

2.3.3. Scenario 3

After the Notice to Proceed (start milestone), activity A can start (FS). Activity B will start one day after activity A has started (SS + 1). Activities C and E will start after activity B finishes (FS). Additionally, activities C and E must finish at the same time (FF). Activity D will start two days after activity C is finished (FS + 2). Activity F starts after Activity E finishes (FS). Activity G will start after activity A is finished (FS) and 2 days after activity F is finished (FS + 2). Activity H will start one day after activity A has started (SS + 1). Activity I will start two days after activity H is finished (FS + 2). Activity J starts after activities D and I are finished (FS). The Project Closeout (end milestone) will be completed once activities G and J are finished (FS). After the completion of the schedule with the assigned program, the end date of 4 March 2024 (03/04/2024) will be reached. Figure 3 shows the network diagram created based on the description of the third scenario, Sc3.

2.4. Questionnaires

A Personal Information questionnaire was designed to ask participants to specify their age, gender, bachelor’s degree, and previous experience with scheduling. The statistical information provided in Section 2.1. Participants are based on the data collected from this questionnaire. To gather participant feedback on their experience using visual and tabular scheduling programs, a total of five questionnaires were designed. Three questionnaires were tailored for each individual scheduling program, referred to as Individual Platform Reflection questionnaires, while two questionnaires were designed to compare the visual with each tabular scheduling program, referred to as Platform Comparison questionnaires.

To define the key functionalities of the program for the purpose of evaluation in these questionnaires, the scheduling process, as well as the overall understanding of the schedule from the perspective of project management, were considered. The steps to create a schedule are (1) define tasks, (2) order tasks based on activity sequencing, (3) establish task relationship, (4) determine task duration, (5) assign resources and budget, (6) process the schedule based on the scheduling method, such as CPM [43]. Since before entering data into the scheduling program, the task order and duration are determined by construction professionals [41,42], steps 2 and 4 were eliminated from consideration. Additionally, this study did not consider resource and budget constraints, eliminating the consideration of step 5. Participants were asked to use the programs in an auto-schedule setting, and the program processed the schedule. Consequently, two key functions of (a) define tasks and (b) define correlations were selected based on the scheduling process. To perform quality control, users check their schedule for mistakes and make changes to the schedule upon spotting them. As a result, two functions of (c) make changes to schedule, and (d) spot mistakes were selected. Finally, the last selected criterion was (e) understanding the overall schedule. These five key functions were the evaluation factors in all the questionnaires.

2.4.1. Individual Platform Reflection Questionnaires

These questionnaires were given to participants upon completion of scheduling with the program assigned to them. The questionnaires were named Platform Reflection Vis, Platform Reflection Tab1, and Platform Reflection Tab2. Each questionnaire consisted of two sections, multiple choice and written. The first section aimed to gauge users’ ease of use across five key functionalities of a scheduling program. Participants were asked to specify to what extent the assigned program was challenging during the scheduling to (1) define tasks, (2) define correlations, (3) spot mistakes in their schedule, (4) make changes to the schedule, and (5) understand the overall schedule. The options provided in these multiple-choice questions were (a) very easy, (b) easy, (c) medium, (d) challenging, and (e) very challenging. The written section allowed participants to specify features that influenced their scheduling process, including those that expedited schedule creation, caused challenges, and supported overall understanding of the schedule.

2.4.2. Platform Comparison Questionnaires

There were two sets of Platform Comparison questionnaires called “Platform Comparison Vis vs. Tab1” and “Platform Comparison Vis vs. Tab2”. They were given to participants at the end of the experiment, following the completion of scheduling tasks with all three programs and subsequent feedback submission through the respective Individual Platform Reflection questionnaires. Each of these questionnaires comprised two sections, multiple choice and written, that were designed to gather insights into participants’ preferences and experience with the scheduling platforms. The first question in the multiple-choice section aimed to identify the platform that provided the most assistance across five key functionalities of a scheduling program. Participants were asked to specify the program that helped them the most during the scheduling to (1) define tasks, (2) define correlations, (3) spot mistakes in their schedule, (4) make changes to their schedule, and (5) understand the overall schedule. Subsequently, the second question sought to ascertain participants’ platform preferences in terms of (6) speed: creating schedule faster and (7) intuitiveness for schedule creation. The third question sought to determine which platform participants deemed more trustworthy in the event of conflicting outcomes between the two programs for the same schedule. For this purpose, participants were asked if they created the same schedule with both of the programs and ended up with two different end dates, which program would they trust more to not make a mistake during scheduling. These three questions gave participants the option to select either of the two programs they compared or state that both programs were the same based on the functions questioned. The first question in the written section asked participants to specify the reason behind their platform selection for trustworthiness based on their response to the last question of the multiple-choice section. The following two questions asked them to specify the features of each platform that they preferred over the other by writing them down in their own words.

2.5. Procedure

Before the start of the experiment, a sample scenario similar to the experiment scenarios was provided to the participants along with the instructions to allow them practice and familiarize themselves with the assignment. Each participant was then randomly assigned to one of the three programs in a way that one-third of the participants started with Vis, one-third with Tab1, and the rest with Tab2. Upon assignment of the scheduling program, participants were asked to log in to the respective program and the university’s web-based learning management system, Canvas [44]. The experiment was designed as a quiz on Canvas that collected the submissions of the students and measured the duration it took them to finish each scenario. All participants started with the first scenario, Sc1. Once logged in and ready, the research team worked with each participant individually to ensure the accuracy of the time measured. Each participant was given an access code to enter into Canvas to start the quiz and was immediately provided with the scenario on paper to start the scheduling task. Upon completion, participants submitted the schedule in PDF format on Canvas, which recorded the exact scheduling time for the respective program. After the completion of Sc1, participants were randomly assigned to another program to complete Sc2, and then Sc3 with the remaining program. Upon each Canvas submission, participants were given the Individual Platform Reflection questionnaire. Upon completion of all three scenarios and questionnaire submissions, participants were given the two Platform Comparison questionnaires. Figure 4 shows the experiment procedure visually.

3. Results

The results of the data analysis are presented in three sections of (1) individual platform reflection, (2) platform comparison, and (3) time spent to create a schedule.

3.1. Individual Platform Reflection

This section presents the analysis of the data collected from the three Individual Platform Reflection questionnaires for Tab1, Tab2, and Vis. The feedback of the participants for the level of difficulty they faced during scheduling based on the five key functions is depicted in

Table 1. Participant responses for five key functionalities of Tab1.

Function	Level of Difficulty
	Very Easy	Easy	Medium	Challenging	Very Challenging
Define tasks	43.6%	30.9%	20.0%	5.5%	0.0%
Define correlations	30.9%	41.8%	20.0%	7.3%	0.0%
Spot the mistakes in the schedule	22.2%	31.5%	38.9%	5.6%	1.9%
Make changes to the schedule	29.1%	38.2%	29.1%	3.6%	0.0%
Understand the overall schedule	32.7%	45.5%	18.2%	3.6%	0.0%
Average	31.7%	37.6%	25.2%	5.1%	0.4%

for Tab1, in

Table 2. Participant responses for five key functionalities of Tab2.

Function	Level of Difficulty
	Very Easy	Easy	Medium	Challenging	Very Challenging
Define tasks	33.3%	42.6%	18.5%	3.7%	1.9%
Define correlations	20.0%	47.3%	21.8%	9.1%	1.8%
Spot the mistakes in the schedule	9.3%	35.2%	35.2%	16.7%	3.7%
Make changes to the schedule	23.6%	36.4%	30.9%	7.3%	1.8%
Understand the overall schedule	18.5%	51.9%	22.2%	5.6%	1.9%
Average	20.9%	42.7%	25.7%	8.5%	2.2%

for Tab2, and in

Table 3. Participant responses for five key functionalities of Vis.

Function	Level of Difficulty
	Very Easy	Easy	Medium	Challenging	Very Challenging
Define tasks	69.1%	25.5%	5.5%	0.0%	0.0%
Define correlations	63.6%	23.6%	10.9%	1.8%	0.0%
Spot the mistakes in the schedule	52.7%	27.3%	14.5%	3.6%	1.8%
Make changes to the schedule	61.8%	29.1%	9.1%	0.0%	0.0%
Understand the overall schedule	66.7%	25.9%	7.4%	0.0%	0.0%
Average	62.8%	26.3%	9.5%	1.1%	0.4%

for Vis. On average, 89.1% of the participants found Vis to be very easy or easy, while this number was 69.3% for Tab1 and 63.6% for Tab2. The average percentage of participants who found the level of difficulty medium for five functions was 10.0% for Vis, compared to 25.2% for Tab1 and 25.7% for Tab2. On average, 1.5% of the participants found Vis1 to be challenging or very challenging, whereas 5.5% did so for Tab1 and 10.7% for Tab2.

To quantify the results and ease the comparison between the platforms, a numerical value was assigned to each level of difficulty: Very Easy = 5, Easy = 4, Medium = 3, Challenging = 2, and Very Challenging = 1. Since there were three programs to compare, to validate the results statistically, a one-way ANOVA was conducted on the data to determine if the differences in the responses towards the scheduling programs were statistically significant [45]. The significance level was set at 5% to test if p < 0.05 [46]. This test was performed based on the responses to each of the five function questions in the Individual Platform Reflection questionnaire. The results indicated that for all five functions, the difference in average response for at least one scheduling program was statistically significant. The results of this test are presented in

Table 4. p-Values of one-way ANOVA and paired t-test.

Statistical Test	One-Way ANOVA	Paired t-Test
	p-Value	p (T <= t) Two-Tailed
Function	Vis–Tab1–Tab2	Vis–Tab1	Vis–Tab2	Tab2–Tab1
Define tasks	0.0005526	0.0045510	0.0003446	0.4729784
Define correlations	0.0000495	0.0009638	0.0001855	0.2518465
Spot the mistakes in the schedule	0.0000024	0.0013244	0.0000099	0.0697631
Make changes to the schedule	0.0000038	0.0000665	0.0000282	0.2417500
Understand the overall schedule	0.0000019	0.0003081	0.0000111	0.1208020
Average	0.0000003	0.0000876	0.0000041	0.1516894

and

Table 5. Test results of one-way ANOVA and paired t-test.

Statistical Test	One-Way ANOVA	Paired t-Test
Function	Vis–Tab1–Tab2	Vis–Tab1	Vis–Tab2	Tab2–Tab1
Define tasks	At least one µ is different	µ Vis > µ Tab1	µ Vis > µ Tab2	µ Tab2 ≈ µ Tab1
Define correlations	At least one µ is different	µ Vis > µ Tab1	µ Vis > µ Tab2	µ Tab2 ≈ µ Tab1
Spot the mistakes in the schedule	At least one µ is different	µ Vis > µ Tab1	µ Vis > µ Tab2	µ Tab2 ≈ µ Tab1
Make changes to the schedule	At least one µ is different	µ Vis > µ Tab1	µ Vis > µ Tab2	µ Tab2 ≈ µ Tab1
Understand the overall schedule	At least one µ is different	µ Vis > µ Tab1	µ Vis > µ Tab2	µ Tab2 ≈ µ Tab1
Average	At least one µ is different	µ Vis > µ Tab1	µ Vis > µ Tab2	µ Tab2 ≈ µ Tab1

.

To compare the average response on platforms between each pair, a two-tailed paired t-test was conducted since the feedback for each of the three platforms was from the same sample pool [47]. The significance level was set at 5% to test if p < 0.05. The paired t-test was performed for responses to each of the five function questions in the Individual Platform Reflection questionnaire. The average response of Vis, µ Vis, Tab1, µ Tab1, and Tab2, µ Tab2 were compared against each other: Vis vs. Tab1, Vis vs. Tab2, and Tab2 vs. Tab1. The results are presented in Table 4 and Table 5. The one-way ANOVA and paired t-test results showed that for each of the five functions, the average response for Vis was significantly greater than those for Tab1 and Tab2. This indicates that more participants reported Vis to be easier to use than Tab1 and Tab2. Additionally, there was no statistically significant difference between the average feedback received for Tab1 and Tab2 across all five functions.

It was also investigated whether previous experience influenced participant responses regarding the level of difficulty for the five key functions. The participant responses are detailed in

Table 6. Participant responses for five key functionalities with and without prior experience of Tab1.

Level of Difficulty	Define Tasks		Define Correlations		Spot the Mistakes in the Schedule		Make Changes to the Schedule		Understand the Overall Schedule
	Exp	No Exp	Exp	No Exp	Exp	No Exp	Exp	No Exp	Exp	No Exp
Very Easy	41.7%	44.2%	25.0%	32.6%	27.3%	20.9%	25.0%	30.2%	33.3%	32.6%
Easy	50.0%	25.6%	58.3%	37.2%	27.3%	32.6%	41.7%	37.2%	50.0%	44.2%
Medium	8.3%	23.3%	8.3%	23.3%	36.4%	39.5%	33.3%	27.9%	16.7%	18.6%
Challenging	0.0%	7.0%	8.3%	7.0%	9.1%	4.7%	0.0%	4.7%	0.0%	4.7%
Very Challenging	0.0%	0.0%	0.0%	0.0%	0.0%	2.3%	0.0%	0.0%	0.0%	0.0%

,

Table 7. Participant responses for five key functionalities with and without prior experience of Tab2.

Level of Difficulty	Define Tasks		Define Correlations		Spot the Mistakes in the Schedule		Make Changes to the Schedule		Understand the Overall Schedule
	Exp	No Exp	Exp	No Exp	Exp	No Exp	Exp	No Exp	Exp	No Exp
Very Easy	55.6%	28.9%	33.3%	17.4%	22.2%	6.7%	22.2%	23.9%	22.2%	17.8%
Easy	33.3%	44.4%	22.2%	52.2%	11.1%	40.0%	44.4%	34.8%	55.6%	51.1%
Medium	11.1%	20.0%	22.2%	21.7%	44.4%	33.3%	11.1%	34.8%	22.2%	22.2%
Challenging	0.0%	4.4%	22.2%	6.5%	22.2%	15.6%	22.2%	4.3%	0.0%	6.7%
Very Challenging	0.0%	2.2%	0.0%	2.2%	0.0%	4.4%	0.0%	2.2%	0.0%	2.2%

and

Table 8. Participant responses for five key functionalities with and without prior experience of Vis.

Level of Difficulty	Define Tasks		Define Correlations		Spot the Mistakes in the Schedule		Make Changes to the Schedule		Understand the Overall Schedule
	Exp	No Exp	Exp	No Exp	Exp	No Exp	Exp	No Exp	Exp	No Exp
Very Easy	100.0%	64.6%	85.7%	60.4%	71.4%	50.0%	85.7%	58.3%	85.7%	63.8%
Easy	0.0%	29.2%	14.3%	25.0%	14.3%	29.2%	0.0%	33.3%	14.3%	27.7%
Medium	0.0%	6.3%	0.0%	12.5%	0.0%	16.7%	14.3%	8.3%	0.0%	8.5%
Challenging	0.0%	0.0%	0.0%	2.1%	14.3%	2.1%	0.0%	0.0%	0.0%	0.0%
Very Challenging	0.0%	0.0%	0.0%	0.0%	0.0%	2.1%	0.0%	0.0%	0.0%	0.0%

for Tab1, Tab2, and Vis, considering their previous experience. The responses are graphically represented in Figure 5, Figure 6 and Figure 7 for Tab1, Tab2, and Vis.

To statistically validate the difference in responses of participants with and without prior experience, a two-sample t-test was conducted [48]. The significance level, α, was set at 5% to test if p < α. The mean response of the participants with experience, µ Exp, was compared against the mean response of the participants without experience, µ No Exp, for each of the key five functions. The results of the t-tests and the two-tailed p-value are detailed in

Table 9. The t-test results for participant responses with and without prior experience of Tab1.

Function	t-Test Results
	p (T <= t) Two-Tailed	Result	Conclusion
Define tasks	0.283	p > α	µ Exp ≈ µ No Exp
Define correlations	0.872	p > α	µ Exp ≈ µ No Exp
Spot the mistakes in the schedule	0.962	p > α	µ Exp ≈ µ No Exp
Make changes to the schedule	0.960	p > α	µ Exp ≈ µ No Exp
Understand the overall schedule	0.628	p > α	µ Exp ≈ µ No Exp
Average	0.696	p > α	µ Exp ≈ µ No Exp

for Tab1, in

Table 10. The t-test results for participant responses with and without prior experience of Tab2.

Function	t-Test Results
	p (T <= t) Two-Tailed	Result	Conclusion
Define tasks	0.078	p > α	µ Exp ≈ µ No Exp
Define correlations	0.831	p > α	µ Exp ≈ µ No Exp
Spot the mistakes in the schedule	0.901	p > α	µ Exp ≈ µ No Exp
Make changes to the schedule	0.859	p > α	µ Exp ≈ µ No Exp
Understand the overall schedule	0.391	p > α	µ Exp ≈ µ No Exp
Average	0.670	p > α	µ Exp ≈ µ No Exp

for Tab2, and in

Table 11. The t-test results for participant responses with and without prior experience of Vis.

Function	t-Test Results
	p (T <= t) Two-Tailed	Result	Conclusion
Define tasks	0.00001	p < α	µ Exp ≠ µ No Exp	µ Exp > µ No Exp
Define correlations	0.03703	p < α	µ Exp ≠ µ No Exp	µ Exp > µ No Exp
Spot the mistakes in the schedule	0.67105	p > α	µ Exp ≈ µ No Exp
Make changes to the schedule	0.49932	p < α	µ Exp ≠ µ No Exp	µ Exp > µ No Exp
Understand the overall schedule	0.08984	p > α	µ Exp ≈ µ No Exp
Average	0.17223	p > α	µ Exp ≈ µ No Exp

for Vis.

Based on the results, it can be concluded that for Tab1 and Tab2, there is no statistically significant difference between the responses of participants with and without experience. Participants with prior experience with visual scheduling programs have rated Vis higher than participants without experience for functions of defining tasks, defining correlation, and making changes in the schedule. There is no statistically significant difference between the responses of participants with and without experience for functions such as spotting the mistakes in the schedule and understanding the overall schedule.

3.1.1. Platform Features That Supported Scheduling

This section presents the results of the written section of participant feedback on the features that helped them create the schedule faster and understand the overall schedule. The results represented in this section do not include responses of none or N/A or those that were left with no response. Non-relevant responses were also excluded from the analysis. It should be noted that some participants specified multiple features that were all considered in the analysis.

Platform Features That Contributed to Faster Scheduling

In this subsection, we present the results based on the participant feedback on the feature that helped them create the schedule faster. The responses of the participants for the three programs were grouped under seven categories: (1) user interface, (2) task creation, (3) correlation assignment, (4) task information, (5) auto-schedule, (6) network diagram, and (7) Gantt chart. The user interface was an umbrella term used by participants referring to different features in the interface to create the schedule, visualize it, etc., while some participants specifically specified schedule creation features that supported them in task creation or relationship assignment, as well as schedule visualization features such as Gantt chart and network diagram as separate features. The task information category referred to the participants’ efforts to look up activity information for purposes such as checking its accuracy. Auto-schedule was the required setting for all three programs in this study that was mentioned in some of the comments. The results are detailed in

Table 12. Percentages of participants reporting features that helped to schedule faster.

Category	Program
	Tab1	Tab2	Vis
User interface	20.0%	12.7%	43.6%
Task creation	12.7%	16.4%	25.5%
Correlation assignment	29.1%	43.6%	18.2%
Task information	14.5%	12.7%	0.0%
Auto-schedule	9.1%	1.8%	7.3%
Gantt chart	10.9%	1.8%	0.0%
Network diagram	1.8%	5.5%	5.5%

, showing the percentage of participants reporting features that helped them to schedule faster.

The overall user interface was reported by a considerable number of the participants for Vis, which was higher than that of tabular scheduling programs, which represented different features for schedule creation and visualization. Participants liked the simple visual interface of Vis and the tabular interface of Tab1 and Tab2, which was similar to the Microsoft Excel [49] program. Looking into schedule creation features, ease of task creation was the second highly reported reason by Vis users that helped them perform scheduling faster by creating shapes, while they reported ease of assigning relationships compared to other features of tabular scheduling as the main reason for helping them schedule faster in Tab1 and Tab2 in a tabular format. Looking up task information and viewing the Gantt chart were reported only for tabular scheduling programs as the tabular interface allowed users to check for information in the table, and the Gantt chart was presented next to the table. Checking the network diagram was reported by a few, even for Vis. Some examples of participant feedback are provided in

Table 13. Feedback examples for features that helped to schedule faster.

Category	Program	Examples of Participant Feedback
User interface	Tab1	It is similar to Excel, which I am familiar with.
	Tab2	Easy interface
	Vis	Easy and simple interface
Task creation	Tab1	Defining the tasks and duration faster
	Tab2	I was able to quickly define the activities.
	Vis	Creating the activities in the visual form
Correlation assignment	Tab1	Assigning successors and predecessors was much faster.
	Tab2	Assigning the correlation is very easy.
	Vis	Creating the relationships between the tasks is fairly simple and straight forward.
Task information	Tab1	Easy access to properties
	Tab2	The detailed chart
	Vis	N/A
Auto-schedule	Tab1	Auto-updating finish date as tasks are added
	Tab2	Auto schedule
	Vis	Automatically schedule is created.
Gantt chart	Tab1	It was easy to view the table side by side with the Gantt chart.
	Tab2	Gantt chart
	Vis	N/A
Network diagram	Tab1	Activity network diagram
	Tab2	Network diagram
	Vis	Visualizing network diagram

.

Platform Features That Contributed to Understanding the Overall Schedule

This subsection presents the results based on the participant feedback on the features that contributed to understanding the overall schedule. The written responses of the participants were grouped into four categories: (1) user interface, (2) task information, (3) Gantt chart, and (4) network diagram. The results are detailed in

Table 14. Percentages of participants reporting features that helped understand the overall schedule.

Category	Program
	Tab1	Tab2	Vis
User interface	18.2%	14.5%	27.3%
Task information	9.1%	14.5%	1.8%
Gantt chart	41.8%	38.2%	25.5%
Network diagram	5.5%	3.6%	16.4%

, showing the percentage of participants reporting features that helped them understand the overall schedule.

While interpreting the results, it should be noted that the user interface in tabular scheduling programs showed a table and Gantt chart, while it was only a network diagram in Vis. The results showed that the Gantt chart was the most reported tool that helped participants understand the overall schedule in tabular scheduling programs. Considering the combination of user interface and network diagram categories for Vis, it can be interpreted that the network diagram was the most helpful tool reported in the visual scheduling program. Nearly one-fourth of the participants using Vis stated the Gantt chart in another interface supported understanding the schedule. On the other hand, very few had checked the network diagram in another interface while using the tabular scheduling programs. This shows the importance of the Gantt chart in supporting users in understanding the overall schedule. Users also found checking the task information to be helpful. Some examples of participant feedback are provided in

Table 15. Feedback examples for features that helped in understanding the overall schedule.

Category	Program	Examples of Participant Feedback
User Interface	Tab1	It is simple and easy to understand.
	Tab2	Interface is good and user-friendly.
	Vis	The interface is very approachable.
Task information	Tab1	The details bar, which makes seeing activity relationships easier
	Tab2	The details tab helps to easily toggle between activity information.
	Vis	Check activity types
Gantt chart	Tab1	Visualization of the Gantt chart with task details
	Tab2	The Gantt chart creation as you create the activities
	Vis	The final Gantt chart
Network diagram	Tab1	Network diagram
	Tab2	Check activity diagram
	Vis	View the network diagram

.

3.1.2. Platform Features That Caused Challenges during Scheduling

This section presents the results based on the participant feedback on the features that caused challenges during scheduling. The written responses of the participants were grouped into four categories: (1) user interface, (2) task creation, (3) correlation assignment, and (4) spotting mistakes. The results are detailed in

Table 16. Percentages of participants reporting challenging features in programs.

Category	Program
	Tab1	Tab2	Vis
User interface	18.2%	20.0%	3.6%
Task creation	5.5%	0.0%	20.0%
Correlation assignment	32.7%	21.8%	18.2%
Spotting mistakes	1.8%	18.2%	1.8%

, showing the percentage of participants reporting features that caused challenges during scheduling.

The distribution of the challenges reported was different for each program. For Tab1, the user interface and correlation assignment were the top two challenges users faced. For Tab2, besides the user interface and correlation assignment, challenges with spotting mistakes was also mentioned by users. The two main challenges participants faced in scheduling with Vis were task creation and correlation assignment. The reasoning behind that was explained as activity shapes overlapping or hiding behind each other during task creation, as well as drawing links between the task shapes when they were not close to each other. Some examples of participant feedback are provided in

Table 17. Feedback examples for features that caused challenges during scheduling.

Category	Program	Examples of Participant Feedback
User interface	Tab1	Knowing where everything is in the software could be the hardest part.
	Tab2	Navigating the overall system can be chaotic.
	Vis	Sometimes it makes us confused.
Task creation	Tab1	Defining tasks
	Tab2	N/A
	Vis	New activity sometimes creates an activity that hides behind an act.
Correlation assignment	Tab1	There was a lot of tiresome clicking when assigning predecessors to activities.
	Tab2	To assign a relationship between the activity is complex.
	Vis	Placing the activities in accordance with the activity relationships
Spotting mistakes	Tab1	Spotting mistakes
	Tab2	When there is an error, it is a little difficult to pinpoint the error in the schedule.
	Vis	Spotting the errors

.

3.2. Platform Comparison

This section presents the results of the analysis based on the data collected from the Platform Comparison questionnaires, comparing Vis versus Tab1 and Vis versus Tab2.

3.2.1. Comparison of Vis vs. Tab1

In this section, the results of the analysis of the Platform Comparison questionnaires that determined the most preferred platform between Vis and Tab1 for various scheduling functions and factors are presented. First, the results of the multiple-choice questions are discussed, and then the responses to the written sections are presented. The results do not include the written responses of none or N/A or those that were left with no response. Non-relevant responses were also excluded from the analysis. It should be noted that some participants specified multiple features that were all considered in the analysis.

Platform Preference (Vis vs. Tab1) for Five Key Functions

The preferences of participants between the platforms to perform five key functions are detailed in

Table 18. Platform preference (Vis vs. Tab1) for five key functionalities.

Function	Platform Preference (%)
	Vis	Same	Tab1
Define tasks	56.4%	30.9%	12.7%
Define correlations	60.0%	30.9%	9.1%
Spot the mistakes in the schedule	60.0%	21.8%	18.2%
Make changes to the schedule	54.5%	36.4%	9.1%
Understand the overall schedule	50.9%	32.7%	16.4%
Average	56.4%	30.5%	13.1%

in percentages and visually presented in Figure 8 with bar charts. The results show that more than half of the participants, ranging from 50.8% to 60.0%, with an average of 56.4%, preferred Vis over Tab1 for all five key functions. The number of participants preferring Tab1 over Vis was relatively low, ranging from 9.1% to 18.2%, with an average of 13.1%. A moderate number of the participants, ranging from 21.8% to 36.4%, with an average of 30.5%, found both programs similar in performing the five key functions.

Platform Preference (Vis vs. Tab1) for Speed and Intuitiveness

The second query in the questionnaire identified the favored platform regarding speed and intuitiveness. The inclinations of participants towards each platform are outlined in

Table 19. Platform preference (Vis vs. Tab1) for speed and intuitiveness.

Factors	Platform Preference (%)
	Vis	Same	Tab1
Faster to create a schedule	80.0%	9.1%	10.9%
More intuitive to create a schedule	63.6%	14.5%	21.8%

and visually presented in Figure 9 with bar charts.

In terms of speed, Vis emerged as the favorite program, with 44 participants, 80.0%, indicating a preference for this platform. In total, seven participants, 12.7%, had prior experience of working with visual scheduling programs similar to Vis, from which five participants, 9.1%, opted for Vis and two participants, 3.6%, opted Tab1. This suggests that most participants found Vis to be faster in creating schedules compared to Tab1. Conversely, only 10.9% of participants favored Tab1 for its speed in schedule creation, while from the total pool of 55 participants, 12 participants, 21.8%, had prior experience of working with Tab1. From these 12 participants, only 2 participants, 3.6%, preferred Tab1 over Vis, and the remaining 10 participants, 18.2%, opted for Vis.

In terms of intuitiveness, 35 participants, 63.6%, indicated a preference for Vis. In total, seven participants, 12.7%, had prior experience of working with visual scheduling programs similar to Vis, from which four participants, 7.3%, opted for Vis and one participant, 1.8%, opted for Tab1, while the remaining two participants, 3.6%, found both platforms equal. In contrast, 12 participants, 21.8%, favored Tab1 for its intuitiveness in schedule creation, while from the total pool of 55 participants, 12 participants, 21.8%, had prior experience of working with Tab1. From these 12 participants, only 2 participants, 3.6%, preferred Tab1 over Vis, and the remaining 10 participants,18.3%, opted for Vis.

Platform Preference (Vis vs. Tab1) for Trustworthiness

The participant responses regarding platform trustworthiness in the event of conflicting outcomes using both programs reveal that the majority, 69.1%, reported the use of the visual scheduling method can reduce the possibility of making mistakes. In total, 12.7% of the participants had prior experience working with a scheduling program like Vis, from which 9.1% of the participants opted for Vis. Conversely, only 21.8% of participants favored Tab1 over Vis. From the total pool of participants, 21.8% of the participants had prior experience of working with Tab1, from which only 1.8% of the participants preferred Tab1 over Vis. The remaining participants, 9.1%, reported both programs performing equally. The results are visually presented in Figure 10 as a pie chart.

The participant responses to the reasoning behind their choice of platform for trustworthiness were categorized under (1) user interface and (2) spotting mistakes, as detailed in

Table 20. Participant feedback on the reason to choose the platform (Vis vs. Tab1) for trustworthiness.

Category	Vis	Tab1
User interface	73.7%	58.3%
Spotting mistakes	23.7%	16.7%

. The percentages reported in Table 20 are among those who chose Vis or Tab1.

The most reported reason for trustworthiness was the user interface, where participants indicated a user-friendly interface could reduce the possibility of making mistakes and consequently result in a schedule that is more trustable. In this analysis, the responses related to visual or tabular methods were grouped under the user interface category as well. The second reason was the ease of spotting mistakes that allowed users to check their schedule for accuracy. Some examples of participant comments are shown in

Table 21. Feedback examples for the platform trustworthiness.

Category	Program	Examples of Participant Feedback
User interface	Vis	Vis is much simpler. Interface is easy to understand and work. We have to just make the network diagram which is easier than entering the data manually or in linear form.
	Tab1	Is more understandable and easy to work on it. It was easier for me to mix up activities in Vis, so they could overlap. The activities are always in the same place for Tab1.
Spotting errors	Vis	It is very easy to check the mistakes we made and to rectify that.
	Tab1	As it shows the Gantt chart, I can refer to and find my mistake.

.

Preferred Features—Vis vs. Tab1

This section summarizes the participant feedback on features they preferred in each program over the other. The responses were grouped under four categories, (1) user interface, (2) task creation, (3) correlation assignment, and (4) spotting mistakes, for Vis, as detailed in

Table 22. Preferred Vis features over Tab1 specified by participants.

Category	Participants (%)
User interface	29.2%
Task creation	25.0%
Correlation assignment	43.8%
Spotting mistakes	12.5%

with feedback examples presented in

Table 23. Feedback examples on preferred features of Vis over Tab1.

Category	Examples of Participant Feedback
User interface	Vis’s visual interface
Task creation	Adding activity in Vis is much easier than Tab1.
Correlation assignment	Using lines to connect the predecessors to the successors helps visualize the relationship.
Spotting mistakes	Spot the mistakes

, and grouped under five categories, (1) user interface, (2) correlation assignment, (3) task information, (4) Gantt chart, and (5) spotting mistakes, for Tab1, as detailed in

Table 24. Preferred Tab1 features over Vis specified by participants.

Category	Participants (%)
User interface	38.2%
Correlation assignment	17.6%
Task Information	20.6%
Gantt chart	20.6%
Spotting mistakes	5.9%

with feedback examples presented in

Table 25. Feedback examples on preferred features of Tab1 over Vis.

Category	Examples of Participant Feedback
User interface	I like that it was like Excel, so it was easier to manage.
Correlation assignment	Defining tasks as successors and predecessors
Task information	Tab1 provides more details … all the information is available all the time.
Gantt chart	You can see both the Gantt chart and the schedule at the same time, so any updates are seen in real time on both.
Spotting mistakes	Can check errors individually

.

Drawing lines between the tasks for assigning task relationships was the favorite Vis feature reported by the participants over Tab1. The next was the overall visual user interface. Subsequently, they reported the ease of defining tasks with the use of shapes as the third feature. Some also mentioned the ease of spotting their mistakes in the program made them prefer Vis over Tab1. Some examples of participant feedback are shown in Table 23.

For Tab1, the overall user interface where they could define tasks and correlations in a tabular format was favored the most by the users. Viewing the Gantt chart in the interface and looking up for information in the table were the second most reported features. Subsequently, they reported ease of assigning relationships between the tasks. A few also mentioned the ease of spotting their mistakes in the program made them prefer Tab1 over Vis. Some examples of participant feedback are shown in Table 25.

3.2.2. Comparison of Vis vs. Tab2

In this section, the results of the analysis of the Platform Comparison questionnaires that determined the most preferred platform between Vis and Tab2 for various scheduling functions and factors are presented. First, the results of the multiple-choice questions are discussed, and then the responses to the written sections are presented. The results do not include the written responses of none or N/A or those that were left with no response. Non-relevant responses were also excluded from the analysis. It should be noted that some participants specified multiple features that were all considered in the analysis.

Platform Preference (Vis vs. Tab2) for Five Key Functions

The preferences of participants between the platforms to perform the five key functions are detailed in

Table 26. Platform preference (Vis vs. Tab2) for five key functions.

Function	Platform Preference (%)
	Vis	Same	Tab2
Define tasks	63.6%	23.6%	12.7%
Define correlations	69.1%	23.6%	7.3%
Spot the mistakes in the schedule	74.5%	14.5%	10.9%
Make changes to the schedule	72.7%	16.4%	10.9%
Understand the overall schedule	61.8%	20.0%	18.2%
Average	68.4%	19.6%	12.0%

in percentages and visually presented in Figure 11 with bar charts. The results show that the majority of participants, ranging from 61.8% to 74.5%, with an average of 68.4%, preferred Vis over Tab1 for all five key functions. The number of participants preferring Tab1 over Vis was relatively low, ranging from 7.3% to 18.2%, with an average of 12.0%. A moderate number of the participants, ranging from 14.5% to 23.6%, with an average of 19.6%, found both programs similar in performing the five key functions.

Platform Preference (Vis vs. Tab2) for the Speed and Intuitiveness

The second query in the comparative survey, comparing Vis and Tab2, identified the favored platform regarding the speed and ease of schedule creation. The inclinations of participants towards each platform are outlined in

Table 27. Platform preference (Vis vs. Tab2) for the speed and intuitiveness.

Factors	Platform Preference (%)
	Vis	Same	Tab2
Faster to create a schedule	85.5%	7.3%	7.3%
More intuitive to create a schedule	61.8%	9.1%	29.1%

and visually presented in Figure 12.

In terms of speed, Vis emerged as the favorite program, with 47 participants, 85.5%, indicating a preference for this platform. In total, seven participants, 12.7%, had prior experience of working with visual scheduling programs like Vis, from which five participants, 9.1%, opted for Vis and two participants, 3.64%, opted for Tab2. This suggests that most participants found Vis to be faster in creating schedules compared to Tab2. Conversely, only 7.3% of participants favored Tab2 for its speed in schedule creation, while from the total pool of 55 participants, 9 participants, 16.3%, had prior experience of working with Tab2. From these 9 participants, only 2 participants, 3.64%, preferred Tab2 over Vis, and the remaining 7 participants, 12.7%, opted for Vis.

In terms of intuitiveness, Vis emerged as the favorite program, with 34 participants, 61.8%, indicating a preference for this platform. In total, seven participants, 12.7%, had prior experience of working with Visual scheduling programs like Vis, from which four participants, 7.3%, opted for Vis and two participants, 3.64%, opted for Tab2, while the remaining one participant, 1.8%, found both platforms equal. In contrast, 29.1% of participants favored Tab2 for its intuitiveness in schedule creation, while from the total pool of 55 participants, 9 participants, 16.4%, had prior experience of working with Tab2. From these nine participants, only two participants, 3.64%, preferred Tab2 over Vis, five participants, 9.1%, opted for Vis and the remaining two, 3.6%, found both platforms equal.

Platform Preference (Vis vs. Tab2) for Trustworthiness

The participant responses regarding platform trustworthiness in the event of conflicting outcomes using both programs reveal that the majority, 65.5%, reported the use of the visual scheduling method can reduce the possibility of making mistakes. In total, 12.7% of the participants had prior experience working with a scheduling program like Vis, from which 9.1% of the participants opted for Vis. Conversely, only 21.8% of participants favored Tab2 over Vis, while from the total pool of participants, 16.4% of the participants had prior experience of working with Tab2, from which only 5.5% of the participants preferred Tab2 over Vis. The remaining participants, 12.7%, reported both programs performing equally. The results are visually presented in Figure 13 as a pie chart.

The participant responses to the reasoning behind their choice of platform for trustworthiness were grouped under two categories, (1) user interface and (2) spotting errors, as detailed in

Table 28. Participant feedback on the reason to choose the platform (Vis vs. Tab2) for trustworthiness.

Category	Vis	Tab2
User interface	69.4%	58.3%
Spotting mistakes	33.3%	16.7%

. The percentages reported in Table 28 are of those who chose Vis or Tab2.

The most reported reason for trustworthiness was the user interface, where participants indicated a user-friendly interface could reduce the possibility of making mistakes and consequently result in a schedule that is more trustable. The second reason was the ease of spotting mistakes, which allowed users to check their schedules for accuracy. Some examples of participant comments are shown in

Table 29. Feedback examples on the platform trustworthiness.

Category	Program	Examples of Participant Feedback
User interface	Vis	Vis is easy to use, create schedules, and understand.
	Tab2	It was easy to add predecessors, which helped to understand the scheduling better.
Spotting mistakes	Vis	It was very easy to check the mistakes we made and to rectify that.
	Tab2	It can be easily checked in a single go.

.

Preferred Features—Vis vs. Tab2

This section summarizes the participant feedback on features they preferred in each program over the other. The responses were grouped under four categories, (1) user interface, (2) task creation, (3) correlation assignment, and (4) spotting mistakes, for Vis, as detailed in

Table 30. Preferred Vis features over Tab2 specified by participants.

Category	Participants (%)
User interface	32.7%
Task creation	21.8%
Correlation assignment	34.5%
Spotting mistakes	9.1%

with feedback examples presented in

Table 31. Participant feedback examples of preferred features for Vis over Tab2.

Category	Examples of Participant Feedback
User interface	Schedule prepared through visualization
Task creation	Adding activity feature is easier than Tab2.
Correlation assignment	Defining the relationship between activities
Spotting mistakes	Easy to spot mistakes

, and grouped under five categories, (1) user interface, (2) task creation, (3) correlation assignment, (4) Gantt chart, and (5) spotting the mistakes, for Tab2, as detailed in

Table 32. Preferred Tab2 over Vis features specified by participants.

Category	Participants (%)
User interface	21.8%
Task Creation	7.3%
Correlation assignment	20.0%
Gantt Chart	9.1%
Spotting mistakes	1.8%

with feedback examples in

Table 33. Participant feedback examples on preferred features for Tab2 over Vis.

Category	Examples of Participant Feedback
User interface	Simultaneous view of charts and data
Task creation	Adding activities and duration to the schedule
Correlation assignment	I like being able to have the predecessors and successors open to assign tasks right away.
Gantt chart	The Gantt chart
Spotting mistakes	Checking errors individually

.

Drawing lines between the tasks for assigning task relationships was the favorite Vis feature reported by the participants over Tab2. The next was the overall visual user interface. Subsequently, they reported the ease of defining tasks with the use of shapes as the third feature. Some also mentioned the ease of spotting their mistakes in the program made them prefer Vis over Tab1. Some examples of participant feedback are shown in Table 31.

For Tab2, the overall user interface where they could define tasks and correlations in a tabular format was favored the most by the users. Ease of assigning relationships between the tasks was also among the top favorites. Subsequently, viewing the Gantt chart in the interface and ease of task creation were reported by some. One person mentioned the ease of spotting the mistakes in the program. Some examples of participant feedback are shown in Table 33.

3.3. Time Spent to Create a Schedule

The time spent by participants to create the schedule of each scenario with three programs was analyzed quantitively. The statistical data of this time is summarized in

Table 34. Statistical data on time spent on each platform to create a schedule.

Program	Mean (Min)	Median (Min)	Range (Min)	Standard Deviation (Min)
Vis	30.9	27.0	[10, 91]	14.9
Tab1	41.0	37.0	[16, 94]	18.9
Tab2	50.9	48.0	[15, 118]	22.4

, which includes the mean, median, range, and standard deviation.

The average time spent by participants to create a schedule with Vis was 30.9 min. This value was 41.0 min for Tab1 and 50.9 min for Tab2. Statistical analysis was conducted on the duration spent on scheduling with each program, utilizing data from all participants across three scenarios and three programs. A one-way ANOVA with a significance level, α, of 5% was conducted to test if p < α. It resulted in a p-value of 0.0000008, indicating there was a significant difference among the means. This suggested that at least one program’s mean value was statistically different from the others. Subsequent paired t-tests were conducted to compare the means between the platforms. The comparison between Tab1 and Tab2 yielded a two-tailed p-value of 0.006982, indicating the mean scheduling time for Tab2 was significantly greater than for Tab1 (μTab2 > μTab1). Similarly, the comparison between Tab1 and Vis resulted in a two-tailed p-value of 0.00322, showing that the mean scheduling time for Tab1 was significantly greater than for Vis (μTab1 > μVis). In summary, the statistical tests confirmed that the visual scheduling method resulted in the least time required for creating schedules, with the mean duration ranking in the order of μTab2 > μTab1 > μVis. The results of these tests are detailed in

Table 35. Result of statistical analysis on the duration spent on creating a schedule.

Comparison	Test	p-Value	Result	Conclusion
Vis–Tab1–Tab2	One-way ANOVA	0.0000008	p < α	At least one µ is different
Tab2–Tab1	Paired t-test	0.0069820	p < α	µ Tab2 > µ Tab1
Vis–Tab1	Paired t-test	0.0032200	p < α	µ Tab1 > µ Vis

.

The impact of prior experience on the time spent to create a schedule with each program was analyzed. For Vis, experienced users had a mean time of 32.1 min, with 7 participants having prior experience, while users with no experience had a mean time of 30.7 min, with 48 participants with no experience. For Tab1, experienced users had a mean time of 35.1 min, with 12 participants having prior experience, compared to 42.7 min for users with no experience, where 43 participants had no experience. Similarly, for Tab2, experienced users had a mean time of 48.8 min, with 9 participants having prior experience, whereas users with no experience had a mean time of 51.4 min, with 46 participants having no experience. The mean time with all three programs is visually presented in Figure 14 with considerations of prior experience.

To examine whether prior experience impacted the time spent to create a schedule with each program, a two-sample t-test with unequal variance was conducted. This test compared the mean time spent by participants with prior program experience, µ Exp, to the mean time spent by participants without prior experience, µ No Exp. The t-test was conducted at a significance level, α, of 5% to test if p< α. The results are detailed in

Table 36. Results of the analysis on the impact of experience on the time spent to create a schedule.

Platform	Vis		Tab1		Tab2
	Experienced	No Experience	Experienced	No Experience	Experienced	No Experience
Mean	32.1	30.7	35.1	42.7	48.8	51.4
Variance	138.1	237.9	286.1	370.9	438.7	525.8
Observations	7	48	12	43	9	46
p (T <= t) two-tailed	0.782		0.194		0.744
Result	p > α		p > α		p > α
Conclusion	µ Exp ≈ µ No Exp		µ Exp ≈ µ No Exp		µ Exp ≈ µ No Exp

. The results showed that the p-value for all three programs was over 0.05. This means no statistically significant difference between the mean time spent to create a schedule was observed between participants with and without experience. Consequently, it can be concluded that prior experience did not have an impact on the time spent on schedule creation in this study.

4. Discussion

This research study evaluated the efficiency of visual scheduling compared to the current industry practice of tabular scheduling for schedule creation. An experiment was conducted with a visual and two tabular scheduling programs that required participants to create three comparable schedules with all three programs. The time spent to create the schedules was measured, and participant feedback was collected with questionnaires where they reflected on their experience with each program and compared the platforms. The study results showed that the time spent in schedule creation with visual scheduling was significantly shorter than using tabular scheduling programs. This was confirmed by the majority of the participants, who reported visual scheduling as a tool that allowed them to create a schedule faster compared to two tabular scheduling programs. Based on the participant feedback on their experience with each platform, it was statistically proven that they found visual scheduling easier to define tasks, define correlations, spot mistakes in the schedule, make changes to the schedule, and understand the overall schedule, which were the five key scheduling functions evaluated in this study. More than half of the participants preferred visual scheduling over tabular scheduling programs to perform the five key scheduling functions. A considerable number of participants found the visual scheduling platform has the potential for making fewer mistakes. Consequently, they stated they would trust their schedule created with visual scheduling to be more trustable in the event of conflicting outcomes compared to tabular scheduling. They also found it a more intuitive method to create schedules.

The analysis of qualitative data based on the written section of the questionnaires revealed that those participants favoring visual scheduling preferred its visual approach by drawing network diagrams, which gave them a better understanding of the overall schedule. They found it user-friendly to create tasks and define correlations that reduced the possibility of making mistakes. They also found it easier to spot their mistakes by checking the network diagram. On the other hand, participants who preferred tabular scheduling reported that it was easier for them to find the tasks and check their information and correlations in a table, while in the visual scheduling interface, they needed to search for the tasks in the network diagram. They also found the tabular interface easy to understand due to their familiarity with the Microsoft Excel interface. The results revealed the importance of the Gantt chart in supporting the users to understand the overall schedule. Tabular scheduling users also found it useful to spot the mistakes using the Gantt chart that showed them the progress of the schedule creation in real time. While a considerable number of participants preferred the visual scheduling method, they also reported finding it challenging to create a schedule by drawing network diagrams because shapes representing the tasks could overlap or hide behind each other, requiring them to drag the tasks around. Drawing lines between the tasks to define the task relationships could also be challenging when the tasks were not close to each other.

4.1. Limitations

The participants were graduate students with a few years of industry experience and exposure to scheduling practices. This required the research team to simplify the schedule scenarios in this study. Another significant limitation was the time constraint. The study required the research team to train students on the scheduling programs in one academic semester, which allowed them to focus only on three scheduling programs. Consequently, the team evaluated only one of the visual scheduling programs compared to two of the frequently used tabular programs in the industry. Since the study was designed as an in-class activity, experiment duration was limited. This required the research team to design schedule scenarios that had relatively few activities and task relationships to allow participants to finish the tasks on time. Furthermore, the scenarios were designed in a way that they could be compared in a controlled experiment and were not real-world schedules.

4.2. Future Studies

Future research studies need to be conducted with industry professionals to explore the efficiency of visual scheduling programs in day-to-day scheduling practices compared to currently used tabular scheduling programs. A design of a similar experiment is suggested with more complex scenarios that reflect real-world schedules to be conducted with industry schedulers. Furthermore, studies that include the cost and resource constraints are suggested to be conducted to test the efficiency of the visual scheduling method.