RESEARCH REPORT, 2019
DUE: 9 October
WORD LENGTH: 2000 words (excluding Abstract and References)
You are required to write approximately 2000 words to complete a journal article format submission reporting on the study you participated in during Week 2 practical classes. Click on the link to see the study https://sydneypsy.qualtrics.com/jfe/form/SV_4MFiOi80bPFayDX
The Method and Results sections of the report have been written for you, so your tasks are to:
??Write a 150 word Abstract summarizing the method, results and conclusions of the study.
??Write an Introduction that provides the background to the study, background literature, the rationale for the experimental design and specifies hypotheses about the expected results.
??Construct a Figure depicting the delayed recall data from Table 2/Graph the Final Recall results in Table 2
??Write a Discussion section that describes and interprets the obtained results, relates them to the background literature and hypotheses outlined in the Introduction, and considers the implications for theories of memory retrieval.
??List the references cited in your report in appropriate APA format. Better reports will make appropriate and effective use of additional references beyond those provided below.
You will obviously need to read the Method and Results section very carefully to write your report, but you do not need to include them in your submission. Otherwise, your report should follow the American Psychological Association (APA) format and referencing style. Remember that correct acknowledgement of sources and referencing is critical. Any attempt to present another author’s work as your own is plagiarism and is treated seriously. To avoid this, be careful to keep track of the source of the information you collect when preparing your report and do not quote or paraphrase extensively from your sources. Instead, think about the information you have read and integrate it into your report in your own words. Plagiarism software will be used to detect breaches of these requirements.
For further information and guidelines for writing APA style reports see:
American Psychological Association (2001). Publication manual of the American Psychological Association (5th ed.). Washington, DC: American Psychological Association.
(Note: You are expected to read beyond these references)
Lehman, M., Smith, M.A., & Karpicke, J. (2014). Towards an episodic context account of retrieval-based learning: Dissociating retrieval practice and elaboration. Journal of Experimental Psychology: Learning, Memory & Cognition, 40, 1787-1794.
Mayer, R.E. (2012). Information processing. In K. Harris, S. Graham, T. Urdan et al. (Eds). APA Educational Psychology Handbook: Vol. 1: Theories, constructs and critical issues. American Psychological Association. (pp. 85-99)
Schwoebel, J., Depperman, A.K. & Scott, J.L. (2018). Distinct episodic contexts enhance retrieval-based learning. Memory, 26, 1291-1296.
Participants. A total of 373 students participated in the study in partial completion of an undergraduate psychology course. Participants’ data were excluded if they failed to complete all phases of the study or if it was not possible to link their data for the separate testing occasions. A further six participants were removed because they recalled one or fewer translations on the Immediate memory test, suggesting that they had misunderstood the instructions or failed to take the acquisition tasks seriously. Following these exclusions, the final sample consisted of 283 participants (average age 20.96 (SD=3.11); 79 male, 203 female, 1 undefined; 82% native English speakers) approximately equally distributed across the 12 experimental conditions (n= 18-25).
Design. A 3x2x2 between-subjects factorial design compared 12 conditions that resulted from manipulating three factors: Context (No Context vs Same Image vs Distinct Images), the Intervening Task (Restudy vs Retrieval Practice) and Attention to context images (Yes/No).
Context was manipulated by presenting a photographic image above each Swahili-English word pair during acquisition and intervening trials. In the same context condition, each word pair was always presented with the same image, while in the distinct context condition a new randomly generated image was presented on every trial. For the no context condition, a white box, matching the colour of the screen, replaced the ‘image’ presented for each trial.
The intervening tasks were either a Retrieval Practice task in which the Swahili cue was presented below the context image with a blank entry box in which the participant was instructed to type the English translation; or a Restudy task in which both the Swahili word and its English translation were presented. The exposure time for the conditions was equated and the same intervening task was repeated twice, with the items in a different randomised order.
Attention to context was manipulated by the instructions provided before the acquisition and intervening phase. In the Attend Context condition, in addition to learning the word-pair, students were instructed to classify the scene depicted in the image as ‘outdoors’, ‘indoors’ or ‘neither’ by clicking a response box with a mouse during the acquisition trials. Images were also presented during the relevant conditions of the intervening tasks, and referred to in the instructions for the Attend Context conditions, but participants were not required to classify them so that they attention was focused on the intervening task. In the No Attention condition, no mention was made of the images in either the acquisition or interval phase and participants were not required to classify them.
Participants were randomly assigned to one of the 12 conditions of the factorial design using an algorithm that ensured a relatively even distribution across conditions.
Materials. Sixteen Swahili-English word pairs were selected from Nelson and Dunlosky (1994). Ninety-six naturalistic scenes were selected from Konkle, Brady, Alvarez, and Oliva (2010) and randomly-paired with the Swahili-English word-pairs so that the word-pair and images shared no obvious semantic relationships. Stimuli were presented on a standard computer screen in colour with the words presented below.
Procedure. The study was conducted as a tutorial exercise in groups of approximately 20 students, each equipped with a standard desktop computer. Prior to the initial learning phase, participants were instructed that they would be presented with Swahili-English word pairs and that their goal was to learn the English translations of the Swahili words in preparation for a final test which would assess their ability to recall the English translations. They completed a practice trial in which they were presented with an example word pair and participants assigned to the Attend Context conditions practised classifying an image.
The acquisition procedure was adapted from Schwoebel et al. (2018) to achieve a high level of initial learning, but time constraints allowed only two, rather than four, acquisition trials. All participants were first exposed to the 16 Swahili-English word pairs and their respective image for the condition, i.e. either a scene in the distinct and same context conditions, or no image, for 5secs with a 500ms inter- stimulus interval. They then completed two acquisition trials in which they were presented with the Swahili word next to a text-entry box for 10 secs, along with either the same image as in the initial trial or a distinct new image, and instructed to attempt the retrieve the English translation. In the Attend Context conditions, in addition to the stimuli, a multiple-choice question was presented asking if the image was indoors, outdoors, or neither to which participant clicked a response box to indicate their choice. This question was also included in the Attend Context/No Image condition, and participants were instructed to respond ‘Neither’. After this, regardless of their response, all participants were shown the correct translation for 5s to provide feedback, followed by a 500ms interval before automatically progressing to the next word-pair/image. Between the acquisition and intervening phases participants completed a 60 sec distractor task in which they solved two-digit addition problems.
For participants assigned to the retrieval practice intervening task, the Swahili word was presented below a same or distinct image with a blank entry box. Participants were given nine secs to attempt to retrieve the English translation of the Swahili word. The list of 16 items was presented twice in a random order, with a 500 ms interval between items, without any feedback.
In the restudy intervening task, the list of 16 Swahili-English word-pairs was presented twice, below same or distinct images. Each pair/image remained on the screen for nine secs with a 500ms interval before the next item was presented.
To introduce a delay of approximately five minutes before assessing recall, all participants then completed an antisaccade task from the Inquisit web software library. This was followed by the immediate recall task in which the 16 Swahili words were listed in a random order on a form with a text-entry box for each word, without any images. Subjects had two minutes, represented in a descending counter at the top of the page, to complete as many of the English translations as possible. They were able to scroll up and down the form to return to previous words.
After a week’s delay students were presented with an unexpected delayed recall test using an identical format to that used for immediate recall.
To assess the extent of learning of the word pairs that occurred during the acquisition stage, an initial analysis was conducted of recall performance for the English translation at the end of the acquisition phase before allocation to the intervening tasks (see Table 1).
No Image Same Image Distinct Images
Attend Context 0.435 (0.037) 0.647 (0.032) 0.321 (0.029)
No Attention 0.539 (0.037) 0.755 (0.023) 0.333 (0.028)
Table 1: Average proportion (and standard error) of English translations correctly recalled at the end of the acquisition phase.
An overall ANOVA assessing the effects of context and attention yielded significant main effects of both factors. The significant main effect of Attention (F(1,277) = 9.08, p .001) reflected higher average recall in the No Attention condition than the Attend Context condition. To determine the source of the significant main effect of Context (F(2,277) = 71.43, p .00001), simple comparisons between the different context conditions were conducted which showed that average recall of pairs presented in the Same Image condition was significantly higher than pairs presented in either the No Image condition (F(1,277)=45.27, p .0001) or the Distinct Images condition (F(1,277)= 26.22, p .0001). Neither of these differences yielded significant interactions with the Attention factor (both p .05) demonstrating that the effects of context at the end of acquisition were equivalent for the Attend Context and No Attention conditions.
Delayed recall test
To assess the impact of the experimental manipulations on long-term memory, the major analysis focused on the measure of delayed recall. An overall ANOVA was conducted on the delayed recall data to assess the main effects and interactions of the three factors. This analysis revealed significant main effects of all three factors. There was significant main effect of Intervening task (F(1,271)=4.47, p=.04), which occurred because the average proportion of translations recalled was significantly higher for the Retrieval Practice condition (0.429) than the Restudy condition (0.369). There were no significant interactions with Intervening task, so the results reported in Table 2 average the data for the two intervening tasks. There were also significant main effect of the type of Context image (F(2,271)= 10.73, p .001) and the Attention factor (F(1,271)=7.44, p .01). These main effects were qualified by a significant interaction between Context and Attention (F(1,271) = 3.56, p=.03). To illustrate the interaction, the mean recall rates for these conditions are presented in Table 2, and depicted graphically in Figure 1.
Table 2: Average proportion (and standard error) of English translations correctly recalled in the delayed recall test.
To determine the source of the significant interaction between Context and Attention, the interaction between the context factor and simple comparisons of each context condition were conducted.
These comparisons revealed that recall was significantly higher for word pairs presented without images than those presented with either same images, F(1,277) = 18.53, p .001, or distinct images. F(1,277) = 14.25, p .001, and that neither of these effects significantly differed between the Attend Context and No Attention condition (both p .10).
However, there was a significant interaction between Attention condition and the difference between Same Image and Distinct Image contexts, F(1,277) = 6.21, p .01: recall was higher for pairs presented with Distinct images rather than Same images in the Attend Context condition (F(1,277) = 5.13, p .03), but the opposite effect was observed in the No Attention condition, although the difference was not significant (F(1,277) = 1.67, p .10).