General findings from empirical studies of longhand vs laptop note-taking

28

note-receiver group in the reviewing section. This could have resulted in non-organic note-taking performance and prevented personal meaning-making process. Moreover, note takers may spend less time reviewing the notes seriously because they had just created the notes a short period of time before. They may have taken less effort in reviewing the notes and taking the posttest.

Horwitz’s (2017) study is the first to investigate the effects of note-taking

modalities on reading comprehension. Compared to lecture conditions, whether laptop text notes harm learning remains relatively unclear. The limitations of this study are threefold: its small sample size may have led to no significant differences in the result; the short time between reading and reviewing may have harmed the motivation of studying notes; and finally, some conceptual questions that did not require reading but common knowledge could have affected test accuracy.

2.3.3 General findings from empirical studies of longhand vs laptop note-taking.

In both lecture and reading conditions, studies of longhand versus laptop note-taking have typically included an analysis of note content and post-reading

comprehension performance. While there is a greater consensus in the findings of note content, contradictory results have been found in test performance (see Table 2).

Findings will be further elaborated in the following sections.

29 Table 2

Summary of the results of relative studies Study Condition Test

Delay

Equal Equal Laptop notes:

more words

equal equal Laptop notes:

more words and more verbatim overlap

30 2.3.3.1 Analysis of note content.

The quality of notes was usually analyzed based on word counts and verbatim overlap. When analyzing the content of different notes, laptop note-taking resulted in significantly more words than hand-written note-taking. This is because typing is usually faster and less laborious than handwriting. While one hand is used to write, up to ten fingers are used to type. In handwriting, a closed system is formed with a pen held in one’s hand (Garman, 1990). On the contrary, the articulators, with fingers typing on the keyboard, work in parallel when typewriting. For casual adult typists, the average typing speed is 41 words per minute (WPM), whereas handwriting speed is around 22 to 31 WPM (Fort, 2014). On top of that, there is a ceiling for

handwriting speed because when WPM increases, legibility decreases (Mangen &

Velay, 2010).

In one of the pioneering studies targeting the potential differences between longhand and laptop note-taking, learners were assigned to either transcribe, i.e.

record as much as possible, or take organized lecture notes, i.e., write in their own words (Bui et al., 2013). On average, notes taken by laptops contained more units of ideas originated from the lecture. Interestingly, in handwriting, explicit instruction of asking learners to write as much as possible did not result in larger proportion of idea units comparing to the organized notes group. One possible explanation may be the ceiling of handwriting WPM imposed by physical limitation (Mangen & Velay, 2010).

Recent research in the free note-taking of lectures and reading passages evinced similar results, with participants using pen and paper writing fewer words than their laptop counterparts (Horwitz, 2017; Mueller & Oppenheimer, 2014).

31

Moreover, using three-word chunks as the measure, more overlaps between students’ notes and lecture transcript were found in the group of laptop users, which implied that using a laptop may result in more verbatim notes (Mueller &

Oppenheimer, 2014). In a follow-up experiment, where learners were explicitly told not to take verbatim notes, the results replicate findings in the previous experiment (Mueller & Oppenheimer, 2014). By the same token, a study on text note-taking also showed more verbatim overlap between reading passages and typed notes (35.47%) comparing to longhand notes (19.98%) (Horwitz, 2017). The ability to type faster than one can write makes it possible to record more words in a limited timeframe but also implies more verbatim notes. While shallower mental processing is included in taking verbatim notes and may undermine encoding benefits, the influence on learning comprehension are still under debate.

2.3.3.2 Comprehension test performance.

Studies on test performances of note-taking focus mainly on input

comprehension and factual recall (e.g., Bui et al., 2013; Horwitz, 2017; Kirkland, 2016; Mueller & Oppenheimer, 2014). As previously mentioned, multiple levels of representation (i.e., surface structure, text-based and situation model levels) affect comprehension and recall (Dijk & Kintsch, 1983). Different tasks were thus designed to assess learners’ understanding of input (Butler, 2010; Rohre, Taylor, & Sholar, 2010; Wolf, 1993). While many of them explored on lecture comprehension, others dealt with reading comprehension. Methods and results of both kinds of studies will be included below.

32

Comprehension is the ability to process audio or textual input, understand the words as they are presented and link back to learners’ prior knowledge (Vandergrift, 2007; William, 2009). As in Kintsch’s Construction-Integration (CI) Model of text comprehension (1988), scope of understanding is located along a local-to-global continuum. Thus, regarding the effects of note-taking, typical tasks of testing comprehension can roughly be divided into two types: local processing and global processing tasks (Mueller & Oppenheimer, 2014; Peper, & Mayer, 1986).

Local processing tasks include verbatim recognition and factual recall of

keywords and detailed ideas. Both recall and lower level comprehension are measured in these tasks. On the contrary, global processing tasks require higher level

comprehension. The abilities to categorize and link different parts of the material, recognize the main concepts, summarize the text and make inferences are assessed in these tasks (Van Dijk & Kintsch, 1983).

Typical local processing tests are identification and recall of detailed facts. For instance, in Bui et al.’s (2013) study, participants were tested on important and unimportant details with multiple-choice questions. Findings in the first experiment have shown that longhand group and computer group performed equally well in immediate posttest. However, in the delayed posttest in the second experiment, participants who took organized computer notes performed better. While longhand note taking was not explored in this experiment, longhand note takers were known to produce more notes in their own words (Bui et al., 2013). Therefore, it would be worth exploring the comparison of longhand versus laptop natural note taking habits.

In studies with natural conditions where participants can freely take notes, learners

33

using different modalities did not show difference on factual lecture or reading comprehension in 30-minute delayed posttests (Horwitz, 2017; Kirkland, 2016;

Mueller & Oppenheimer, 2014).

On the contrary, the benefits of encoding have been proved to be more helpful when completing global processing tasks in the empirical study of Mueller and Oppenheimer (2014). In the posttest containing multiple-choice and short-answer questions, longhand participants outperformed their laptop counterpart on conceptual and application tasks. However, this superiority wasn’t significant in Kirkland’s (2016) study on lecture comprehension and Horwitz’s (2017) study on reading

comprehension. There was no difference in the performance between two groups with different note-taking modalities. One reason may be that conceptual comprehension was tested in multiple-choice questions in these studies. Another may be that listening and reading are two different information processing systems and that their results could not be directly compared. The generalization of the results from previous studies are still debatable and further research is therefore needed.