Central and Divided Visual Field Presentation of Emotional Images to Measure Hemispheric Differences in Motivated Attention

All methods described here have been approved by the Internal Review Board for human subject research at the University of Kansas, Lawrence, KS.

1. Selecting Participants

1. Use right-handed participants for DVF research. In rare cases (10%), left-handed individuals are lateralized for language processing in the right hemisphere, which would result in scalp-recorded ERP components with non-typical topographical distributions.
2. Have participants complete the Edinburgh Handedness Inventory⁹ to determine strong right-handedness. Scores of eight or higher indicate strong right-handedness.

2. Stimuli

1. Request a research copy of the International Affective Picture System (IAPS)¹⁰ via the Center for the Study of Emotion and Attention's website¹¹. Select stimuli from the IAPS according to the specifications in steps 2.2-2.4. The IAPS comes with an image file for each stimulus and a tab-delimited text file containing the normed ratings of valence and arousal for each image.
   1. Use a spreadsheet program to view the norms and make the stimuli selections. For a complete list of the stimuli selected for O'Hare, Atchley, and Young (2016) see Table 1.
      NOTE: This stimulus set provides norms for the rated valence and arousal of emotional stimuli. The norms for the stimuli were created via participant ratings on the Self-Assessment Mannequin¹⁰. This scale depicts a graphic figure that ranges from a frowning, unhappy figure to a smiling, happy figure for valence, and a relaxed, sleepy figure to an excited, wide-eyed figure for arousal. Valence is rated on a 9-point Likert scale with 1 equaling the most unpleasant (frowning, unhappy figure) and 9 equaling the most pleasant (smiling, happy figure). Arousal is also rated on a 9-point Likert scale with 1 equaling the least arousing (relaxed, sleepy figure) and 9 equaling the most arousing (excited, wide-eyed figure). The components of each image that evoke emotional responses are located centrally in each image.
2. Create three valence groups of images with 60 images in each group: unpleasant, pleasant and neutral, using the norms provided in the IAPS manual¹².
   1. To do this, sort the IAPS images by their average valence rating. Unpleasant stimuli range in average valence rating from 1 to 3.99. Neutral stimuli range in average valence rating from 4 to 6.99. Pleasant stimuli range in average valence rating from 7 to 9. Each valence group must significantly differ from each other in average valence rating with no overlap in their ranges.
   2. Confirm that valence groups significantly differ from each other using independent samples t-tests¹³. Picture complexity across the image groups is not controlled as image complexity has not been found to influence the LPP¹⁴.
3. Within both the unpleasant and pleasant valence stimuli, create high and medium arousal subgroups of 30 images each.
4. Within the neutral valence stimuli, create medium and low arousal subgroups. High arousal subgroups range in average arousal ratings from 4.30 to 8.70 and do not significantly differ from each other on average arousal rating. Medium arousal subgroups range in average arousal ratings from 2.40 to 7.29 and do not significantly differ from each other on average arousal rating. The low arousal subgroup ranges in average arousal ratings from 1.4-5.44.
5. Once the stimuli have been selected, test them (via t-tests)¹³ to ensure the stimuli groups are reliably different.
   NOTE: Each of the arousal subgroups (high, medium, and low) must significantly differ from each other in average arousal rating, but arousal subgroups within a valence group do not significantly differ from each other in valence. This allows for examination of 1) valence effects alone, 2) arousal effects alone, and 3) interactive effects between valence and arousal.
6. Using an image-editing software program, resize the final stimulus images to ensure that they will be presented at 17.06 horizontal and 10.85 vertical degrees of visual angle on the stimulus-presentation monitor.
   1. Calculate the visual angle (V) using the formula, V = 2arctan(S/2D)¹⁵, where S = the height or width of a visual object, and D = the distance from the viewer's pupil to the visual object. The size of the stimulus images will depend upon the distance between the participant's pupils and the stimulus-presentation monitor (D).
7. Create mask stimuli for backward masking of the image stimuli. Mask stimuli consist of an array of backward slashes (i.e., "") that match the spatial dimensions of the images. Create a textbox that has the same dimensions in pixels as the image stimuli in an image-editing software program. Enter backward slashes into the textbox until they fill the entire space without changing the dimensions specified. Save this textbox as an image to create the mask stimulus.
8. For the DVF paradigm, create the image-presentation slides to be loaded into the stimulus presentation software.
   1. In an image-editing software program, center a fixation mark ("+") in the middle of the image. Place your first stimulus image centered vertically with its right edge 3° of visual angle to the left of the fixation mark.
   2. Create a brown rectangle with the same dimensions as the stimulus image and place it also centered vertically with its left edge 3-degrees of visual angle to the right of the fixation mark. Save this arrangement as the left visual field presentation of this stimulus image.
   3. Switch the location of the stimulus image and the brown rectangle and save this arrangement as the right visual field presentation of this stimulus image. Do this for all stimulus images (Figure 1).
9. For the DVF paradigm, create the mask-presentation slides to be loaded into the stimulus presentation software in the same manner as was done for the image-presentation slides. Place the mask image on both sides of the fixation mark with both inner edges 3-degrees of visual angle from the fixation mark (Figure 2). Save this arrangement as the mask stimulus for the DVF paradigm.

3. Experimental Equipment

1. Use Silver-Silver chloride (Ag-AgCl) active-electrodes or other EEG electrodes to record EEG from scalp positions according to the international 10-20 system¹⁶. Position one additional electrode above and another below the right eye to record vertical eye movements.
2. Use EEG acquisition software for data acquisition with a sampling rate of 250-500 Hz, depending upon equipment specifications. For a detailed consideration of EEG acquisition parameters see Luck (2014)¹⁷_.
3. Present stimuli via an stimulus-presentation software package¹⁸ on a computer with a mirrored 24-inch stimulus-presentation liquid crystal display monitor (1,920 x 1,200 resolution) that is in a separate, electrically-shielded, and sound attenuated room. Place only the mirrored monitor within the shielded room, while keeping the computer out of the experimental room reduces electrical noise. Sound attenuation reduces the occurrence of auditory-evoked potentials in the EEG data. The stimulus-presentation software package must allow for users to set the presentation durations and screen locations of stimuli.

4. Preparing the Participant

1. Have participants complete informed, written consent prior to providing any data.
2. Have participants complete a demographic survey to provide sex, age, handedness, native language, vision, and neurological history. Collect sex and age for reporting in final study dissemination. Use all other demographic information to determine if the participant meets the criteria for inclusion in the study: right-handed (assessed via the Edinburgh Handedness Inventory), native English speaker collected via self-report (or native to the language used in the study instructions), normal or correct-to-normal vision, and no history of neurological trauma.
3. Apply EEG electrodes onto the participant. Any EEG montage that covers occipital-parietal regions of the scalp is appropriate for recording the LPP response.
4. Seat participants in a dark, electrically-shielded, sound-attenuated room. Use a chin rest to stabilize the head and minimize movement. Position the chin rest the correct distance away from the stimulus-presentation monitor to maintain the D variable used in the visual angle calculations. Place a keyboard (or response box) in front of the participant for response collection via their right hand.
5. Check the data signal to ensure that all channel impedances are less than 50 kiloohms¹⁷.
6. Instruct participants to passively view the image stimuli without shifting their eyes away from the center of the screen. Display a fixation mark ("+") at the center of the screen to help participants fixate¹⁷. Instruct the participant that there will be a recognition quiz following each block of images, so it is important that they pay attention. Each participant only completes the CVF or the DVF paradigm, creating a between-subjects design.
   NOTE: Both CVF and DVF paradigms can be conducted on the same participant to create a within-subjects design. If this is done, counterbalance the order of the two paradigms to control for any familiarity effects with the stimuli.

5. Central Visual Field (CVF) Paradigm

NOTE: In the CVF paradigm, randomly present image stimuli at the center of the screen. Each trial consists of a 500 ms central fixation ("+") followed by a 150 ms presentation of the stimulus, followed by a backward mask that varies randomly in presentation duration between 2,000-4,000 ms. Jittered presentation duration for the mask serves to reduce any anticipatory ERP responses to the onset of the next trial²⁰.

1. To specify the presentation durations and stimuli locations create separate presentation slides for the fixation, the stimulus images, and the mask stimulus in your stimulus-presentation software.
   1. For the presentation of the fixation mark, specify the presentation of the plus sign symbol ("+") centered both vertically and horizontally and set the duration to 500 ms. This can be done through the properties for this slide.
   2. For the presentation of the stimulus, enter the image file names for the stimuli into a matrix or list object.
   3. On the image-presentation slide, place an image object centered both vertically and horizontally, and link this object to the list of image file names to load the image stimuli. Set the matrix or list object with the image file names to randomly select from the list without replacement of already selected stimuli. Set the duration of the image-presentation slide to 150 ms.
   4. For the mask-presentation slide, again place an image object centered both vertically and horizontally. This object can be directly linked to the mask image file by entering the file name in its properties. Set the duration of the mask-presentation slide to randomly vary between 2,000-4,000 ms.
2. Present image stimuli in four experimental blocks of 45 trials each (180 trials total). Each block has an equal number of stimuli from the valence/arousal conditions. This can be accomplished by creating four separate matrices or list objects with the image file names, each containing 7-8 images from each valence-arousal group (e.g., in list 1 there can be 7 high-arousing unpleasant images and in list 2 there can be 8 high-arousing unpleasant images). Participants passively view the image stimuli on each trial.
3. Following each block, give a 10-item recognition test to ensure participants pay attention during the passive viewing portion of the study. Display six items on the recognition test from the preceding block and four items that are new. Select these six items so that they represented all categories of valence and arousal. Have participants respond via key press using their right hand indicating which stimuli they previously viewed.

6. Divided Visual Field (DVF) Paradigm

NOTE: The DVF paradigm is identical to the CVF paradigm, including the size of the image stimuli, except present each image stimulus laterally, to the left or right of the fixation mark using the image-presentation slides created in step 2.7 (see Figure 3)⁴.

1. Present each image once in the left visual field and once in the right visual field. Present all stimuli in a fully randomized order.
2. As each stimulus is presented twice, double the number (8) of experimental blocks and recognition tests for a total of 360 trials.
3. Pair each image stimulus with the simultaneous presentation of a solid brown rectangle identical in stimulus dimensions on the opposite side of the fixation. This is done to reduce reflexive saccades to the stimulus. Additionally, the presentation duration of 150 ms is shorter than most saccade latencies²¹, meaning that if the participant shifts his/her eyes to the stimulus, it will be masked before the participant can fixate on it²².
4. Present each stimulus and its paired brown rectangle with their inside edge 3° of visual angle from the fixation. This is done to assure that the stimuli fall entirely within the regions of the retinas that are processed by only one hemisphere⁴.
5. Backward mask both the stimulus and the brown rectangle using the same criteria and procedure as was done in the CVF paradigm²⁰.

7. Data Analysis

1. Remove any participant scoring less than 50 percent (chance) on the recognition test from the data, as it cannot be assured that he/she was paying attention to the stimuli.
2. Preprocess and analyze EEG data using an EEG software package²³. Filter data offline with a continuous 0.01-30 Hz bandpass, mark data fluctuating 200 μv within a 100 ms time window as bad, correct eye blink artifacts using an average template generated from each individual participant, manually remove horizontal eye shifts from the data following visual inspection, and rereference data using the common average rereference^24,25.
3. Compute epochs of 1,000 ms after the onset of the stimuli according to a 200 ms pre-stimulus baseline²⁶.
4. Use visual inspection of the waveforms and the ERP literature to determine the topography of the LPP²⁷. In this study, the LPP centered on channel CPz. In this case, average channels CPz, Pz, Cz, CP1 and CP2 together to create a representation of the LPP.
5. In the DVF data, conduct a laterality analysis comparing the LPP amplitude across left occipital-parietal channels and right occipital-parietal channels to ensure that the DVF presentations did not shift the typical topography of the LPP component. Conduct paired-samples t-tests between channel pairs CP1 and CP2, CP3 and CP4, C1 and C2, C3 and C4, P1 and P2, and P3 and P4 respectively to ensure that they do not significantly differ from each other on average amplitude.
6. As the LPP is a long, sustained component, extract two different LPP epochs: early (400-700 ms post-stimulus onset) and late (700-1,000 ms post-stimulus onset).
7. Analyze the CVF LPP data via a 3 (Valence: unpleasant, pleasant, and neutral) by 2 (Epoch: early and late) within-groups analysis of variance (ANOVA) to ensure that the typical LPP effect of emotional stimuli generating larger LPP responses than neutral stimuli is present. This analysis is done to affirm that the stimuli were processed normally.
8. To examine the interactive effects of valence and arousal on the LPP conduct a 2 (Valence: unpleasant and pleasant) by 2 (Arousal: high and low) by 2 (Epoch: early and late) within-groups ANOVA on the CVF LPP data.
9. To examine the effects of hemisphere of presentation, conduct the ANOVA specified in section 7.8 with the additional factor of Hemisphere: left and right on the DVF LPP data.