Department of Communication Arts
	Research	Links	Streaming Examples

Contents • Chapter 1 • Chapter 2 • Chapter 3 • Chapter 4 • Chapter 5 • References

CHAPTER V

DISCUSSION AND RECOMMENDATIONS

Introduction

This study explored achievement and beliefs as measures of the effectiveness of streaming video as an augmentation to text and image-based Web-based higher education instructional environment. Demographic data obtained in this research study were gender, age, academic college, undergraduate/graduate status, location of Internet access used for the study, either home or campus computer lab, and prior experience level with the Web. Data were collected through online surveys and forms, and analyzed using SPSS for Windows (Release 8.0). Research questions were investigated that dealt with achievement differences in the presence or absence of streaming video, the relationships between demographic characteristics and beliefs about streaming video, relationships between achievement scores and demographic characteristics, and characteristics of streaming video that could lead to beliefs about streaming video in a Web-based instructional environment.

Summary of the Demographic Data

The sample for the research study comprised 103 students at the University of Montevallo, a public liberal arts university in Alabama, who were enrolled in various Summer 1, 2001, classes. By chance, the distribution of the gender, age, academic college, and undergraduate/graduate status of the sample reflects that of the actual University of Montevallo demographic makeup (Fact Book 1999-2000, n.d.). The sample was 38% male and 62% female, while the population of the University of Montevallo in 1999, the latest year for which data are available, was 32% male and 68% female. The largest age group for the sample was the 18-22 age range, which was 71% of the sample. In 1999, 81% of the University of Montevallo’s population was in the age range 17-22. The rest of the sample’s age ranges also reflected the 1999 University of Montevallo population. One third of the sample was enrolled in the College of Arts and Sciences, with the other three colleges each providing approximately one fifth of the sample. In 1999 approximately one half of the students at the University of Montevallo were enrolled in the College of Arts and Sciences, with the other three colleges enrolling between 15 and 20% each of the population. Finally, 84% of the sample were undergraduates, which mirrors that of the 1999 university population, where 81% of the students were undergraduates. The design of the study originally targeted undergraduates, but by including graduate students, the sample more closely reflected the total population of the university. The makeup of the sample and its parallels to the University of Montevallo population are advantageous. The parallels will allow for limited generalization back to the total population of the University of Montevallo. Generalization will be limited due to the small sample size, the 50% completion rate of the research instruments, and the design of the study in which the characteristics of the study at a single point in time were examined.

A majority (58%) of the participants accessed the treatment material from home. This is advantageous given the low-speed dial up connection most Internet users have at home. While high-speed broadband access is available in many areas, only 14% of the Internet users in North America currently have high-speed Internet connections (Broadband: Europe Remains Behind in Broadband Game, 2001). Therefore, issues relating to streaming video over low bandwidth connections will have to be taken into account for some time to come.

Approximately one quarter of the participants described themselves as having minimal Web experience. One third were self-reported novices, and approximately 38% reported having an intermediate level of experience with the Web. Only three of the participants described themselves as having expert Web experience. The distribution of students’ experience is noteworthy, because streaming video is a relatively new technology. Software required for viewing streaming video clips is in a constant cycle of advancement, and it is likely that users with a novice level of experience or less will be comfortable upgrading or installing software required for use with streaming video. Additionally, users who do not have at least an intermediate level of Web experience may be uncomfortable using the technologies involved in streaming video, or even understand what they are and how to use them. Furthermore, if technical problems arise with streaming video clips, less experienced users may not know how to solve the problem, and could become turned off towards the technology.

A total of 77 students participated from seven classes, but only 51 completed the instrument. More subjects in the streaming video treatment group did not complete the study than did subjects in the no streaming videos treatment group. This attrition may have been due to problems encountered with the streaming video clips. As Chute, Thompson, and Hancock (1999) noted, if students are unhappy with the system, the discomfort may become a distraction that inhibits learning. A number of subjects attempted the instrument twice, but those were weeded out using a code the researcher had asked the subjects to use if they had to repeat the treatment for any reason. Although the small sample size limits the generalizability of the study, the results from this exploratory study can still provide insight into the effectiveness of streaming video in Web-based instruction.

Research Question One

The first research question compared achievement scores of groups either viewing or not viewing streaming video clips. Streaming video clips were used in this study as an enhancement to text and still image Web-based instruction. Mean comparison tests were performed to determine if viewing or not viewing streaming video clips would produce significant differences in achievement scores. As Mahendran and Young (1998) noted, achievement can be related back to effectiveness. Both groups showed significant achievement gains from the pre-test to the post-test. This result is not surprising when considered against the sample demographics. The instrument used Microsoft Access as the subject for the online tutorial. The majority of the courses from which the sample was derived were first- or second-year-level courses. Two of the courses, MIS 161, Introduction to Computer Applications, and ED 335/527, Instructional Technology I, were introductory computer courses, and students would not be expected to be familiar with database concepts, particularly at the beginning of academic term as was the case in this study. The sample demographics, combined with comments made by certain of the subjects during the qualitative interviews, and random comments made by some of the subjects when invited to participate, lead to the conclusion that the majority of the sample had had little to no exposure to database concepts prior to this study. It is possible that the significant achievement gains for both groups from pre-test to post-test were a result of either pre-test or post-test sensitization due to the similar nature of the tests. Therefore, a gain in mean achievement scores across both groups from a score of 5 to a score of 40 was not unexpected. Edwards and Fritz (1997) provide another interpretation of the gains by noting that students perceive strong learning outcomes from Web-based material. In this case, because the students were aware of participating in a Web-based research study, an effect relating to the student’s perception of learning due to the use of Web-based instruction may have occurred.

A comparison of the mean pre-test scores of Group 1 and Group 2 found no significant difference in achievement scores between the groups. Comparing the mean post-test scores of Group 1, no streaming video, to Group 2, streaming video, showed no significant difference in achievement scores. The results of this mean comparison test showed that using streaming video as an enhancement to Web-based instruction is as effective as text and still image Web-based instruction. This supports Jung and Rha’s (2000) conclusion that there is no general learning superiority for one type of technology over another.

Research Question Two

Achievement gains are one measure of effectiveness. The use and perceived effectiveness of streaming video in Web-based instruction as measured by beliefs are another. The second research question asked whether or not demographic characteristics would lead to significantly different beliefs about the use and perceived effectiveness of streaming video as an enhancement to text and image based Web-based instruction. Madachy and Miller (1978) noted that the integration of videotapes and computer-aided instruction had distinct advantages over the use of either medium by itself. Extending this statement to the new technology of streaming video, this research question will examine the use and perceived effectiveness of streaming video when it is integrated with Web-based instruction, which may be thought of as a form of computer-aided instruction. Twelve belief statements measured on a 4-point scale were administered to Group 2, able to view streaming video clips, through an online survey. Chi-Square tests were performed for each demographic versus each belief item. Gender was not found to lead to significantly different beliefs about the use and perceived effectiveness of streaming video in Web-based instruction. Interestingly, the age range of the participant was found to lead to significantly different beliefs on four of the belief items (see Table 9). Younger participants were more likely to agree or strongly agree with the belief statements, while participants over the age of 40 were more likely to disagree or split between disagree and agree with the belief statements. As can be seen in Table 9, the statements deal with understanding and learning from the streaming video clips, while some of the non-significant belief statements deal with concepts such as enjoyment or technical attributes of streaming video. This supports findings by Chun and Plass (1994) who noted that QuickTime video clips were found to be more helpful in a language learning study than text alone. One possibility for this relationship between age and beliefs about streaming video may be due to learning style preferences. Stokes (2001) found that students who preferred visual learning had a significantly lower mean age than students who preferred verbal learning. Stokes described visual learners as preferring charts, diagrams, and pictures, while verbal learners prefer to listen or to read. Streaming video clips allow a user to see moving and still images, which would seem to favor visual learners over verbal learners.

Academic college was also found to lead to a significantly different belief about the attention-holding ability of streaming video clips (see Table 9). Participants enrolled in the College of Arts & Sciences or the College of Fine Arts either agreed or strongly agreed that streaming video clips helped hold their attention. Students enrolled in the College of Business either disagreed or agreed that the clips helped hold their attention, while students either undecided about their academic college or enrolled in the College of Education disagreed that the streaming video clips helped hold their attention. No data were collected that might explain this difference between academic colleges.

Participants’ status as either undergraduate or graduate also showed a significant difference on the attention-holding ability of streaming video clips (see Table 9). The overwhelming majority of undergraduate students tended to agree that streaming video clips helped hold their attention, while all of the graduate students disagreed that the clips helped hold their attention. One possibility for this may be that graduate students tend to be older and, at the University of Montevallo, enrolled in the College of Education. Age has been shown to lead to significantly different beliefs on visual-oriented belief statements, and participants enrolled in the College of Education have been shown to disagree that streaming video clips helped hold their attention. A 1996 study by Wise and Groom also found that various forms of video programs held the interest of most secondary students better than instructor lecture alone. While secondary students are not equivalent to undergraduate students, the finding that video programs held interest better than lecture alone may have implications for the use of streaming video as an enhancement to instruction with students in secondary schools or undergraduate college programs.

The location participants used to access the treatment material, either home or campus computer lab, led to a significant belief about the amount the subjects learned by having the streaming video clips available (see Table 9). All participants who accessed the clips from a campus computer lab either agreed or strongly agreed that they learned more with the streaming video clips than they would have without them. Students who accessed the clips from home were just about equally split between disagreeing and agreeing or strongly agreeing that they learned more with the clips than without them. This is an interesting finding about the use and perceived effectiveness of streaming video in Web-based instruction. Students who accessed the clips from a campus computer lab were able to view the clips over a dedicated high-speed Internet connection, and they all agreed or strongly agreed that they learned more with the clips. Students who accessed the clips from home, however, were almost as likely to disagree as to agree that they learned more with the streaming video clips available. Although no data were gathered about the type of computer and Internet connection participants had at their home, this points out the uncontrolled nature of home access for use in Web-based instruction. While not all campus computer labs will have the latest hardware or software, a glance into most labs will reveal similar machines that are usually maintained by a dedicated support staff and that are likely connected to the Internet via a high-speed connection. Students who access Web-based instructional material from home may have anywhere from an older computer to the latest model, and may be connecting via analog dial-up modem or a high-speed dedicated connection such as cable broadband or digital subscriber line (DSL). An explanation for this result is that some subjects who accessed the treatment materials from home may have experienced problems with the streaming video clips and thus did not feel they learned as much with the streaming video clips as without them. As Hecht and Klass (1999) noted in a report on the use of streaming video to deliver a doctoral level course, problems with the delivery of lectures via streaming video led to students requesting the course instructor to forego streaming video and teach in person. The nature of the Internet access therefore needs to be taken into account when using streaming video in Web-based instruction.

Research Question Three

The third research question explored whether or not demographic characteristics related to differences in achievement scores. One-way ANOVA tests were performed comparing each demographic characteristic with pre-test and post-test scores. Tests were performed for the sample as a whole, and for each treatment group. No significant differences were found between demographic characteristics and differences in achievement for the entire sample of participants. There were also no significant differences found between demographic characteristics and differences in achievement for participants in Group 2, the streaming video group. For subjects in Group 1, no streaming videos available, however, significant differences were found on the pre-test scores for the age, undergraduate/graduate status, and Web experience demographics (see Table 10). As can be seen in Table 10, subjects above 40 years of age scored higher than younger subjects did on the pre-test.

Graduate students in Group 1 also scored higher on the pre-test than did undergraduates. Finally, subjects with an intermediate level of Web experience scored higher on the pre-test than did subjects with either a minimal or novice level of Web experience. These results, while significant, are hard to interpret in light of the lack of significant differences for the sample as a whole, and for the streaming video group, Group 2. Perhaps age, undergraduate/graduate status, and Web experience have exposed students to terms relating to databases, thus allowing those subjects to score higher. The results, however, seem to be inconclusive in light of only one subgroup from the entire sample displaying significant relationships.

Research Question Four

Four random participants, two from each treatment group, were interviewed to determine if there are characteristics of streaming video in a Web-based instructional environment that lead to beliefs about the use and perceived effectiveness of streaming video in a Web-based instructional environment. This qualitative component was included because there has been virtually no previous research on the effectiveness of streaming video, and there is no recognized theory relevant to the topic.

A phenomenological approach to the interview analysis was chosen to examine the theory that streaming video augments static Web-based instruction. The interview transcripts were analyzed, and certain categories of answers and themes emerged.

The first two subjects interviewed were members of Group 1. The subjects were previously identified as Subject 1, an 18-22 year old male enrolled in the College of Fine Arts with intermediate Web experience and who accessed the treatment from home, and Subject 2, an 18-22 year old female with a novice level of Web experience who is undecided about her major, and who accessed the treatment from a campus computer lab. The next two subjects were interviewed from Group 2, identified previously as Subject 3 and Subject 4. Subject 3 was an 18-22 year old male enrolled in the College of Fine Arts who accessed the treatment materials from home and described himself as having expert Web experience. Subject 4 was an 18-22 year old female enrolled in the College of Arts & Sciences who described herself as having intermediate Web experience and who also accessed the treatment materials from home. A number of interesting themes emerged from the questions concerning the streaming video clips, asked only of these two subjects.

Questions Targeted to the Streaming Video Treatment Group

First, a theme of the streaming video clips acting as a learning reinforcement ran through many of the answers given by both subjects. Both subjects felt that the streaming video clips were helpful in learning. Subject 3 noted that he watched clips only when he needed clarification of a topic in the text of the tutorial. Subject 4 noted that she liked the reinforcement the clips offered the most. She stated that hearing the host talk and explain concepts from the accompanying text was a great method of reinforcement for her. Subject 4 went so far as to identify herself as “visually stimulated” and further stated that she would read the text first, and then watch the clips to know if she had interpreted the text correctly. Smith and Ransbottom (2000) found that video can complement text and both subjects stated that they preferred having streaming video clips as an enhancement to a text and static image Web-based tutorial rather than viewing an entire course or tutorial by streaming video. Subject 3 made a point of stating that the streaming video clips enhanced a regular Web tutorial by breaking up long sections of textual content. Both subjects also felt the clips were easy watch and provided an avenue for immediate review of the content and learning. This theme of the streaming video clips acting as a learning reinforcement supports the results of the Chi-Square comparisons of demographic characteristics with beliefs statements. The Chi-Square comparisons showed that younger students tended to believe that the streaming video clips were easy to understand, helpful in learning and that adding streaming video clips to Web pages would make the Web pages better (see Table 8).

A related theme found in the interview statements is the belief that the streaming video clips helped to hold the attention of the subjects. Subject 4 stated the clips helped maintain her attention on the subject matter of the tutorial, and likewise, Subject 3 noted that the use of the clips as part of the tutorial broke up the long sections of text, and made the amount of content in the tutorial less overwhelming. Subject 3 also noted that the clips added an element of motion to the tutorial, which helped to hold his attention, and to make the tutorial seem more interactive. Holding one’s attention is often motivated by stimulating interest, and as the Toronto Board of Education (1972) found, telecast portions of lectures stimulated pupil interest. Streaming video may be related to telecasts since both methods involve the transmission of video from source to receiver. This particular theme is also reflected in the Chi-Square comparisons that showed undergraduate students in the College of Arts & Sciences and the College of Fine Arts believed the clips helped to hold their attention.

The third theme that emerged from the interviews relates to the attributes of the streaming video clips. Both subjects stated that the screen size of the streaming video clips did not affect the way they felt about the clips. This is supported by D. J. Dwyer’s (1985) study of the impact of screen size and resolution on response accuracy and time on a PC board chip locator task. D. J. Dwyer concluded that screen size and resolution had no practical impact on the locator task. The streaming video clips as used in the study were designed to display at 160 pixels by 120 pixels resolution, which is considered to be one quarter of a normal screen size. Subject 3 noted that he is used to a small window, and was able to make out detail in the video window even at the low resolution used for the clips. Subject 4 was also able to make out detail in the screen shots, although she reported that the clips appeared at a full screen resolution, which was puzzling since the clips were not designed to play at that size. This unforeseen occurrence was attributed by the researcher to an uncontrolled condition involving Subject 4’s computer and the streaming video clips. Both subjects stated that they liked the appearance of the clips, and felt that the pacing of the clips and the use and number of Microsoft Access screen shots was favorable.

Questions About Web-Based Instruction in General

The next sets of questions were common to all 4 interview subjects and dealt with Web-based instruction as a whole. Again, a number of interesting themes emerged from their answers. All four subjects felt that they were able to learn from the Web-based tutorial. Subjects 1 and 2, who did not view the streaming video clips, were not as positive about the amount of learning they experienced, as were the subjects who were able to view the streaming video clips. When asked if they would like to take an entire course via the Internet, all except Subject 1 responded positively. Subject 4 reiterated her belief that the streaming video clips acted as reinforcement to the textual content and thus aided their learning. Subject 3 remarked that Web-based instruction is “like having a one-on-one experience with the instructor” while discussing his experiences with the tutorial. This interesting quote from Subject 3 concerning his satisfaction with the Web-based tutorial enhanced with streaming video clips echoes findings by Boverie, Murrell, Lowe, Zittle, Zittle, and Gunawardena (1997) whose data analysis revealed a moderate to high relationship between the social presence of the television instructor, in this case the host of the streaming video clips in conjunction with the text-based tutorial, and overall student satisfaction. Subject 2 noted that a course via the Internet would allow for flexible scheduling and would provide students with an unbiased, consistent course structure. Subject 1, noting a preference for human contact, stated that he prefers regular courses with Web-based study guides and other enhancements to an entire course via the Internet.

An unexpected theme was a preference for the use of streaming video clips in conjunction with a visual learning style that was suggested by three of the four subjects. Subject 1 noted that he would have preferred to see the streaming video enhanced version of the tutorial. He described himself as having a visual learning style, learning more from pictures and demonstrations than from plain text. Subject 1 was aware of the streaming videos from the instructions given by the researcher and felt that he could have learned more from the version of the tutorial enhanced with streaming videos. Subject 3 also felt that he learned more from the tutorial enhanced with streaming video clips, and stated that the streaming video clips added motion and interactivity which helped to hold his interest better than would plain text. Subject 4, who described herself as “visually stimulated” noted that the streaming video clips added a lot to the tutorial, helped focus her attention, and helped her to retain the content she read. Taken together, these statements point to a theme of an interaction of the streaming video clips with a visual learning style.

A dislike for the length of the tutorial was the third theme that emerged from questions regarding Web-based instruction. Subjects were asked what they liked least, or what they would change, and all of the interview participants mentioned the amount of content in the tutorial or the length of the tutorial page as a negative aspect of the Web-based tutorial. The subjects felt it was long and drawn out, and needed to be broken up into smaller sections.

Additional Analyses

Additional analyses of the data were conducted to further explore the effectiveness of streaming video in Web-based instruction. Results from the beliefs survey were correlated with the demographics of Group 2, streaming video clips available, to determine if significant relationships between beliefs and demographic groups exist. A second opportunity for additional qualitative analysis presented itself in the form of a student whom the researcher allowed to take both treatments. The student was then queried about her feelings and beliefs concerning both versions of the Web-based tutorial.

Pearson’s r Correlations for Demographic Characteristics and Beliefs

Pearson’s r correlations were calculated between the demographic characteristics of participants in Group 2, the streaming video group, and the beliefs statements. A significant moderate negative correlation was found (r(25) = -.495, p = .009) between the age of the subject and the belief that streaming video clips helped the subject understand the material. Younger participants tended to believe that the streaming video clips helped them to understand the material, with older participants disagreeing that the clips helped them with the material. Undergraduate/graduate status and the belief that the clips helped hold the subjects attention showed a significant moderate negative correlation (r(25) =     -.673, p = .000), with undergraduate students agreeing that the clips helped hold their attention, while graduate students disagreed with that statement. A significant moderate positive correlation (r(25) = .437, p = .023) was found between the belief that streaming video clips helped hold the subject’s attention and the Web experience level of the subject. As the subject’s Web experience level increased, their tendency to believe the clips helped hold their attention increased. The beliefs concerning attention holding and learning from the streaming video clips are supported by Ellis and Childs’ (1999) study that found videos viewed within a Web page useful for learning.

The pre-test score of the subject showed a significant moderate positive correlation (r(25) = .385, p = .047) with the belief that subjects experienced no problems with the streaming video clips. Interestingly, the higher the subject’s pre-test score, the stronger the subject’s agreement that he or she had experienced no problems with the streaming video clips. Finally, a significant positive moderate correlation was found (r(22) = .497, p = .013) between the post-test score of the participant and the belief that the participant could learn as much from the streaming video clips as from watching television. The data showed that as subjects scored higher on the post-test they tended to agree more with the belief about learning as much from streaming video clips as from television.

Analysis of Interview with Comparison Subject

A fifth interview was conducted after an opportunity arose in the person of a student who was enrolled in two classes in the sample. The researcher decided to let her take the treatment a second time as a member of Group 2 in order to gain insight into the subject’s perceptions and beliefs regarding streaming video and Web-based instruction.

The student is an 18-22 year old female enrolled in the College of Fine Arts who described herself as having minimal Web experience. She first participated as a member of Group 1, unable to view streaming videos. The second time the subject viewed the tutorial she viewed the streaming video version from a campus computer lab. Of the two versions of the tutorial, the subject reported liking the version enhanced with streaming video clips the most. The subject noted that having the streaming video clips is “kind of like having somebody there to point things out.” In a response moments later when the student was asked if the streaming video clips made the subject matter easier or harder to learn, she replied that the video clips made it seem as if she “had someone doing it in front of her.” These are very similar to the statement made by Subject 3 in the previous set of interviews. Subject 3 stated that Web-based instruction is “like having a one-on-one experience with the instructor” while discussing his experiences with the database tutorial. As was previously noted, a 1997 study by Boverie, Murrell, Lowe, Zittle, Zittle, and Gunawardena revealed a moderate to high relationship between the social presence of the television instructor and overall student satisfaction. Perhaps streaming video clips used in conjunction with Web-based instruction encourages these feelings of the social presence of the instructor. If this is so, streaming video clips may prove to be very useful in enhancing student satisfaction with Web-based instruction.

The subject was also asked if the screen size of the streaming video clips affected the way she felt about the streaming video clips. She responded that she was able to understand what was going on in the streaming video clips and that the subtitles added by the researcher helped her to follow the screen action. This supports Strachan’s (1996) supposition that the human visual system can still make sense of a video sequence that has been compressed enough to be either downloaded or streamed over the Internet.

Streaming Video Production Notes

Creating the streaming video clips for this research study proved to be a challenging yet positive process. Although the researcher has considerable experience with traditional television production, the production of streaming video clips as an enhancement for Web-based instruction was a new experience. A review of the literature did not reveal guidance for best practices, so the researcher had to rely on trial and error methods.

Production of streaming video clips for use in Web-based instruction entails answering many questions. Decisions must be made on the size and appearance of the streaming video clip on the computer monitor. Additionally, trade-offs must be made between video quality and the message of the clip. One has a dizzying array of choices when publishing the clips on the Web, from the software used for streaming to the data rate at which the clips are transmitted over the Internet. It is hoped that the experiences gained from this study will be helpful in future instances of streaming video in Web-based instruction.

Video Production Notes

Based on the researcher’s previous television production experience, the decision was made to use the highest quality equipment available to the researcher for the streaming video clip production process. The researcher had access to both VHS and Mini DV format camcorders. A Sony VX-2000 Mini DV camcorder was chosen because it would record video with much higher resolution and color accuracy than possible with a VHS camcorder (Zettl, 2001 pp 235-240). All videos were edited on an iMac DV using Final Cut Pro 1.25 nonlinear editing software. Final Cut Pro 1.25 is a professional quality video editing program. The videos were saved as Windows AVI format files at full screen resolution, 720 x 486 pixels, and scaled down to quarter screen resolution, 160 x 120 pixels using Adobe Premiere 4.1 on a Windows 95 PC. After scaling, clips were encoded for streaming using Real Networks RealProducer 7.0 software on a Windows 95 PC.

An issue arose when the researcher noticed that the scaling ratio meant approximately 10% of the horizontal pixels were unaccounted for when the video was scaled from 720 to 160 pixels wide. An attempt was made to scale using whole integers, down to a size of 180 x 120 pixels, but the image looked distorted. The decision was made to trust the scaling of the Adobe Premiere software that dealt with the pixels in question and produced a natural looking picture. It is unknown if these missing pixels had an effect on the actual or perceived usefulness of the streaming video clips.

The host was recorded in a television production studio at the University of Montevallo, with great attention given to lighting and audio quality. All screen close-ups were recorded later using a laptop computer set to 800 x 600 pixel resolution. A laptop computer was used because CRT displays produced dark rolling bands in the picture, while the LCD screen on the laptop did not. Television studio lighting was used to produce a pleasing picture of the host with sufficient contrast and intensity levels. A dedicated clip-on microphone was used for the host to capture the highest quality audio possible while she read the script from the teleprompter. Reeves and Nash, as cited in King, Harnar, and Mayall (1999) noted that the fidelity of the audio is the most critical component of media with regard to getting the message across.

As part of the decision to use best practice video production techniques to produce the highest quality streaming video clips possible, the researcher decided to use screen close ups and subtitles rather than screen shots of the entire Microsoft Access window exclusively. Close up screen shots were included at various points for maximum clarity of the concept presented in the video clip. Subtitles were used to guide the subject’s attention and emphasize certain elements of the video window. Timing of the close-ups and subtitles was determined by the focus of the host’s dialogue and the pace of the video clip. If the host referred to a specific menu, a close up shot and a subtitle would often be used for emphasis (see Figure 2). Full view screen shots, close ups and subtitles would be alternated to avoid overuse of any one technique, and to avoid fast and repetitive transitions between screen shots. Fast and repetitive transitions were found to produce unacceptable images when streaming.

Figure 2. Microsoft Access screen close up with subtitles.

Processing of Clips for Streaming

The decision was made to have the clips play at quarter screen resolution, 160 x 120 pixels. Although screen size and resolution were limited by this decision, it was felt that this would be a good test of the effectiveness of streaming video in Web-based instruction. Additionally, as Ohanion (1998, chapter 3) noted, if the intent of the image could be preserved the tendency of the human eye to blend contiguous areas and infer meaning would produce a recognizable image. Figure 3 is an example of the typical screen shot that students viewed. Subjects reported being able to discern details in the image. Furthermore, trials showed that clip playback would be smoothest over analog modems at this resolution.

Clips were encoded using RealProducer 7.0 for playback over 56K modems and T1 lines, and preference was given first to audio quality then to sharpness of the video images. The production process used for the streaming video clips was arrived at by trial and error. The final process used was felt by the researcher and by pilot study evaluators as the best available for clip playback at quarter screen resolution.

During initial attempts to produce streaming video clips for this study, it was discovered that the video clips appeared very dark when streamed over the Internet. This proved to be the biggest hurdle to finding acceptable clip quality. The clip image quality issue appeared to be related to the use of Final Cut Pro 1.25 to edit the clips. A workaround was developed that required saving the clips as AVI files with the brightness slightly increased over a normal image. When clips were then scaled down using Adobe Premiere 4.1, the brightness was slightly increased again.

Figure 3. Screen shot of the data table as seen in the RealPlayer window.

Finally, when clips were processed in RealProducer 7.0, a two-pass variable bit rate method was used. This complicated process produced images that displayed at normal brightness on Windows PC computers. It was later discovered, however, that on Macintosh computers, the clips would playback with a slightly washed-out appearance (see Figure 4). The reason for this playback discrepancy was never discovered, as all clips were played on both types of computers using Microsoft Internet Explorer 5.0 and RealPlayer 8.0. It appears that the type of computer used for playback influences the appearance of the clips. The subject who viewed both versions of the Web-based tutorial viewed the clips on a Macintosh computer, however, and did not mention the washed out images.

 Figure 4. Typical screen shot of the host in the RealPlayer window.

Limitations

This research project was subject to a number of known limitations. One limitation is that the study included only students enrolled at a single institution in the southeast region of the United States. Another limitation is that the location and type of computer used by each subject to participate in this study was uncontrolled. This may have biased beliefs about the effectiveness of streaming video due to the possibility of different computers having displayed the content of the study differently from the way in which the researcher had intended.

Data transmission over the Internet, which is out of the control of the researcher due to the existing design and infrastructure of the Internet itself, may have introduced unintended biases. The streaming video used in the study is designed to be displayed on the subject’s screen at a predetermined size, which, combined with the uncontrolled type of computer used by the subjects and differences in video display due to the nature of data transmission over the Internet, may have caused differences in the appearance of the streaming video.

The small sample size used in this study was another limitation. A number of the mean comparison tests were unusable due to the small sample size. It was discovered during data collection that the instructor for two of the larger enrollment courses in the sample withdrew his offer of credit for participation, and as a result, most students in his courses did not participate. Additionally, some students in the streaming video treatment group apparently had problems with the treatment materials, as more repeat attempts were noted and weeded out than in the no streaming videos treatment group. It is not known what impact this had on the results, or if all of the repeat attempts were removed from the data analysis. Graduate students were also included in the data analysis through a weakness in the design of the experiment. The original design did not call for graduate students, but it was not discovered until data collection had begun that the process for removing graduate students from the sample would not work.

Conclusions

This study examined the effectiveness of streaming video in Web-based instruction. Effectiveness was measured by achievement gains and by beliefs about the perceived use and effectiveness of streaming video clips as an augmentation to a Web-based tutorial. Assessing the effectiveness of streaming video in Web-based instruction is important because if the use of streaming video in Web-based instruction grows as rapidly as the Internet seems to be growing, then at some point its instructional effectiveness will be called into question.

Research Question One

Using streaming video as an augmentation to a Web-based tutorial was not found to produce significant gains in achievement over a text and still image Web-based tutorial. In the experiment, both groups made significant achievement gains from pre-test to post-test. This supports Jung and Rha’s (2000) meta-analysis of the effectiveness of online education where they found no general learning superiority for one type of technology over another. However, a correlation was found between post-test scores of individuals in the streaming video treatment group and the belief that they could learn as much from streaming video clips as they could from watching television. While television is not normally considered a component of Web-based instruction, the literature showed examples of instruction being delivered by television-based technologies such as videotape, over the air broadcast, satellite transmission and others. If students can truly learn as much by watching streaming video clips as they can by watching television, perhaps streaming video will be the technology of choice in the future for delivering video instruction. As Pescatore (2000) noted, streaming video is now considered an acceptable medium for a wide variety of uses even though it does not meet broadcast television standards. Future advances in technology may elevate streaming video to the level of broadcast television.

Research Question Two

 The second research question examined whether or not demographic characteristics of the streaming video treatment group would lead to significantly different beliefs about the use and perceived effectiveness of streaming video in Web-based instruction. This question is important because if demographic characteristics do play a role in beliefs about the perceived effectiveness of streaming video, then instruction using streaming video could be targeted more effectively. Age range was found to lead to significantly different beliefs on 4 of the 12 beliefs statements. Younger students tended to agree more with statements concerning the usefulness of streaming video clips in learning while older students did not as a whole agree with those statements. This is noteworthy because while students at the University of Montevallo are typically under the age of 25, not all institutions of higher education will have demographics that match those of the University of Montevallo. The demographics academic college and undergraduate/graduate status were found to lead to significant differences on beliefs about the attention-holding ability of streaming video clips. Undergraduates rather than graduate students tended to believe that streaming video clips helped hold their attention. Students enrolled in the College of Arts and Science, the College of Business and the College of Fine Arts also tended to believe that streaming video clips helped hold their attention more than did students undecided on major or enrolled in the College of Education. No data are available on the personality characteristics of the students that might explain why students in the College of Education tended to disagree that streaming video clips helped hold their attention. While instruction should not focus exclusively on holding the attention of the student, a condition where the presentation becomes more important than the content, being able to hold the student’s attention is a useful feature. Using streaming video as an augmentation to Web-based instruction may be a useful way to make Web-based instruction more interactive and dynamic. As Hart, Hart, and Benavides (1992) concluded, distance learning lectures must be dynamic to be effective. Web-based instruction as a form of distance learning should therefore be dynamic.

The significant differences in beliefs found in the demographics of age range, academic college and undergraduate/graduate status, when taken together, provide a strong argument for the usefulness of streaming video as an augmentation to Web-based instruction. Younger students enrolled in the colleges of Arts and Sciences, Business, and Fine Arts were found to believe that streaming video is helpful in learning, enhances Web pages, and holds their attention. Further study may reveal exactly why streaming video is effective for these particular demographic characteristics, which could lead to methods to enhance the effectiveness of streaming video for other demographic groups.

The location used to access the treatment materials, either home or campus computer lab, led to significantly different beliefs regarding the perceived amount of learning with the streaming video clips. All of the subjects who accessed the material from a campus computer lab agreed or strongly agreed that they learned more with the clips than they would have without them, while students who accessed the material from home were split about equally between agreeing and disagreeing that they learned more with the streaming video clips than without them. Since streaming video relies on the Internet for delivery, this seems to indicate that care must be taken when students will access course material from home. Until every home user has high-speed Internet access, streaming video may work best as an augmentation to Web-based instruction rather than the only method for content delivery. As Chu and Schramm (1975) found, instructional television, which could include streaming video, works best when it is an integral part of instruction.

Research Question Three

Data analysis for the third research question was performed to determine if demographic characteristics related to differences in achievement. Tests on the sample as a whole and on the streaming video treatment group found no significant differences. Tests performed on the Group 1, no streaming videos available found significant differences on pretest scores for age range, undergraduate/graduate status, and Web experience. Because these results were found for only one subgroup, no convincing conclusions can be drawn from the significant relationships found.

Significant correlations were found between certain belief statements and age range, undergraduate/graduate status, Web experience level, and pre-test and post-test scores. Age range was moderately negatively correlated with the belief that the clips helped the participants understand the material. This supports the results of research question two, where significant relationships were found for the age range of the subject and the perceived effectiveness of streaming video. It is interesting to note that undergraduate/graduate status and Web experience level produced significant correlations for the same belief concerning attention holding. The age of the participant correlating with amount of understanding gained by the use of the streaming video clips and the correlation between undergraduate/graduate status and attention holding seem to indicate a relationship between undergraduates, who are typically younger, and amount of learning. Perhaps there is a trait of streaming video that encourages attention holding in younger students.

Problems with the streaming video clips as indicated by belief item 4 were moderately positively correlated with the subject’s pre-test scores, with scores rising as subjects more strongly agreed that they experienced no problems. This correlation may be due to chance, or students may have experienced problems and, when they started the treatment over, may not have bothered to answer the questions on the pre-test honestly.

Post-test scores were moderately positively correlated with the belief that the subjects could learn as much from watching streaming video as they could from watching television. As scores rose, subjects were more likely to agree with that belief statement. This correlation has significance for the use of streaming video in Web-based instruction. If subjects can indeed learn as much from watching streaming video as they can from watching television, then streaming video is a viable technology to use in distance learning.

Research Question Four

The fourth research question used qualitative methodology to seek characteristics of streaming video that would lead to beliefs about the effectiveness of streaming video in a Web-based instructional environment. A number of noteworthy themes emerged from the interview analyses.

The idea that the use of streaming video clips acts as a learning reinforcement when used as an augmentation to a text-based tutorial emerged from both subjects interviewed from the streaming video treatment group and the subject who viewed both treatment versions. Along with the idea of reinforcement, the subjects also stated that the streaming video clips enhanced the tutorial. Additionally, the subjects thought the streaming videos were useful in learning, which was also emerged from the results of the beliefs survey. These statements and ideas are significant as it implies that the students found streaming video clips used as an augmentation to Web-based instruction to be useful and effectiveness. This is supported by previous studies by the Toronto Board of Education (1972), Chun and Plass (1994), Ellis and Childs (1999), and Smith and Ransbottom (2000) on the use of video in instruction that found that video is helpful in learning. While this result should not be taken to imply that all Web-based instruction should be augmented with streaming video, it does support the argument that streaming video is useful in Web-based instruction.

One subject explicitly stated the streaming video clips helped hold her attention. Subjects who accessed the streaming video clips from a campus computer lab also agreed that the clips helped hold their attention. No information was gathered on the reasons behind this, but it may be related to the idea of the streaming video clips acting as a reinforcement. This is another important characteristic that emerged from the interviews. One subject stated that the streaming video clips added an element of motion and interactivity. It is possible that this is the source of the attention-holding ability of streaming video clips. As Hart, Hart, and Benavides (1992) noted, distance learning lectures must be dynamic and interactive. Further, Wise and Groom (1996) found that videos helped hold attention. One use for streaming video clips in Web-based instruction then, may be as a tool to hold the attention or stimulate the interest of the student. However, this should not be taken to mean that streaming video clips should be used exclusively or in the place of quality content, just that streaming video clips might be best when they are used as an integral part of Web-based instruction.

Unexpectedly, the interview subjects talked about feeling as though the instructor were present. The subjects noted that the streaming video clips generated this feeling. Perhaps streaming video clips used in conjunction with Web-based instruction encourages these feelings of the social presence of the instructor. If this is so, streaming video clips may prove to be very useful in enhancing student satisfaction with Web-based instruction. This is reflected in a 1997 study by Boverie, Murrell, Lowe, Zittle, Zittle, and Gunawardena, who found a correlation between satisfaction and the social presence of the instructor.

Two of the subjects also believed the streaming video clips suited their learning style. Learning style preferences were not investigated as part of this study, but it is intriguing that this topic emerged from the interviews. The two subjects described themselves as visually oriented learners, and it may be that streaming video clips are better suited for one style of learner than another, or may be beneficial in general. Further study is warranted on this topic.

Interestingly, the small screen size and low resolution of the video due to the amount of compression required for streaming did not seem to negatively impact perceptions of the streaming video clips. The interview subjects all had positive feelings towards the clips, and the results of the beliefs survey show the subjects were also positively disposed towards the clips. The interview subjects reported being able to make out detail on the screen, and to learn from the clips. This supports D. J. Dwyer’s (1985) results, although it was thought that the smaller size and lower resolution of streaming video would negatively impact beliefs about the effectiveness of streaming video in Web-based instruction. Apparently, the ability of the human eye to discern detail (Strachan, 1996) is robust enough to make use of streaming video clips, at least for the sample used in the study. Additionally, a previous study found that the fidelity of the audio is the most critical component of media with regard to getting the message across (Reeves & Nass, as cited in King, Harnar, & Mayall, 1999). The fidelity of the audio coupled with the ability of the human eye to discern detail may lead to the belief that streaming video is useful and effective.

The positive beliefs found towards the usefulness of the streaming video clip support the production method used to create the clips. While this one study is not definitive and does not address best practices for streaming video clip creation, it is encouraging that traditional television production techniques may be applied to streaming video. Although not every video produced for another medium may translate well or effectively to streaming video, if care is taken to maximize video and audio quality, and to include techniques such as subtitles, which enhance the message of the clip, then streaming video may be a viable transmission method for a wide variety of video programs and content.

Recommendations

This was an exploratory study designed to examine the effectiveness of streaming video in Web-based instruction. Streaming video clips were not found to produce significant differences in achievement between groups either viewing or not viewing the clips. While streaming video was perceived to be effective by the subjects, a number of questions remain regarding the application of streaming video in Web-based instruction.

Streaming video suiting a visual learning style was a theme that emerged from the study. It is recommended that investigators explore the interaction of streaming video clips with learning style preferences. Will one or more learning styles benefit from streaming video in Web-based instruction more than others? Returning to the debate between Web-based instruction and traditional classroom instruction, would a Web-based course enhanced with streaming video clips lead to significant gains in achievement over a traditional lecture-style course? The Web would seem to offer some advantages for different learning styles, since it allows multiple content delivery modes such as text, audio, and video to act as reinforcements to one another if the student needs it. Along similar lines, it is recommended that future studies compare groups of students who have intentionally chosen one instructional delivery method over another.  This study used random assignment of students in on-campus classes.  Are there characteristics of students who choose completely Web-based instruction over classroom lectures, characteristics that the use of streaming video could better serve?  Further studies could examine the impact of adding streaming video elements to existing Web-based courses as well as adding streaming video elements to existing lecture-style courses. These characteristics could include learning style preferences, but other characteristics may arise as well.

Additionally, it is recommended that further studies explore the interaction of age and learning style preferences with the perceived or actual effectiveness of streaming video. Since age of the participant was previously shown to lead to significantly different beliefs about the effectiveness of streaming video, and because a learning style preference theme arose from the interview analysis, further studies could determine if age and learning style interact to impact the effectiveness of streaming video in Web-based instruction.

One subject noted that he would prefer streaming video enhanced Web-based instruction over a book because he felt the streaming video added an element of interactivity. Further study regarding the interactivity of streaming video is warranted. Streaming video technologies allow for “hotspots” to be placed in the video window that the user can click on to trigger a Web browser event. The technology also allows for self-timed events to occur (RealNetworks, n.d. b). Other technologies such as SMIL and Macromedia’s Flash allow the creation of streaming video lectures that closely resemble a Microsoft PowerPoint presentation (Spotlight on multimedia, n.d.). It is recommended that future studies examine these technologies to determine if they are superior to streaming plain video. Another question that may be explored is if there is a way to take advantage of this technology to better serve the student, or to better serve various learning styles? Future studies may find that streaming video is an excellent way to promote interactive learning as described by Tapscott (as cited in Stokes, 2001).

It is recommended that best practices be investigated specifically for the production of streaming video clips. Although traditional television production techniques seemed to work in this study, that may not always be the case. Questions that need to be addressed include determining the optimal pacing of streaming video clips, the optimal size for objects to appear on screen, and the optimal brightness of the clips. These elements all interact one way on a traditional television screen. Do they interact entirely differently when viewed over the Internet? Also, should streaming video clips be limited to a certain length? If length of time plays a role in the effectiveness of streaming video clips, is that related to technical issues with the streaming software or to something else?   Because the production of streaming video clips proved to be a fairly involved process, it is recommended that future studies explore various methods of production to determine the most efficient method from a developer’s standpoint.  If streaming video is to be used in instruction, timeliness will be an important factor in its success.  Students, as well, will want instructional materials in a timely manner and institutions will want to make good use of their faculty and staff’s time.

The streaming video clips used in this study employed fairly detailed images of Microsoft Access. In 1969 F. M. Dwyer Jr. found that students learned more from abstract line drawings and detailed drawings than from photographs and concluded that the instructional effectiveness of visual illustrations is a function of the type of learning task they are designed to complement. It is therefore recommended that further study be given to the amount of detail used in streaming video images to be able to best match the style of video with the learning task.

Technical issues raise a number of questions and recommendations for further research. This study utilized RealNetworks’ software, but there are a number of other methods for delivering video over the Internet. Although not all of them are true streaming video, it is recommended that further studies examine the different technologies to see if one method is better suited than another, or if one method, such as QuickTime, is suited for a certain application while another method, such as RealPlayer, is better suited for a different application. Another recommendation is to study the effect of bandwidth on the effectiveness of streaming video clips. Until every student’s home is equipped with a high-speed connection, bandwidth will be an issue for the use of streaming video in Web-based instruction. Bandwidth concerns impact issues such as video window size, amount of compression, and the rationing of bandwidth between audio, video motion, and video detail. It is recommended that future studies explore these topics to determine best practices for the delivery of streaming video clips.

Aside from technical issues relating to streaming video, it is recommended that further study be performed to determine what kind of content works best with streaming video. The subjects regarded the content matter used in this study as being complex. Would a different kind of content, one geared towards the affective domain, be better suited to use with streaming video than either the cognitive or psychomotor domains (Bloom’s Taxonomic Classification of Objectives, n.d.)

Due to the limited sample size of this study, it is recommended that further study be given to the relationship between age and perceived effectiveness of streaming video. It is also recommended that this study be replicated, and that future studies compare streaming video-only tutorials with streaming video clips used as an augmentation to a text-based tutorial.  

While the results of this study suggest that streaming video is perceived as effective in Web-based instruction, many questions about this new technology remain. Streaming video will probably not be the one technology that rewrites the future of instruction, but it may prove to be a very useful tool. If for no other reason than students perceive it to be useful as the subjects in this study did, then institutions and instructors should not be afraid of implementing some form of streaming video.

 Home • Research • Links • Examples • Contents • Chapter 1 • Chapter 2 • Chapter 3 • Chapter 4 • Chapter 5 • References