experiment im not sure how to start off. plz help
First Assignment: Psychophysics and Weber's Pick a sensory modality for which it is fairly easy for you to precisely manipulate quantity over a wide range. Some suggestions are weight, sweetness, time and saltiness. It will be helpful to use Excel for this assignment. If you do not have Excel, you can get it for free (https://umdearborn.edu/offices/information-technology- services/software-computer-store). Note that there is a sample spreadsheet (in Canvas Files, called Sample WeberData.xls, that you can use as a starting point (but of course you must fully modify it to fit your assignment). You may not use the same base levels as in the sample spreadsheet. 1. Indicate which modality and thing you have chosen. For example you can indicate Weight using grains of rice or Weight using Cheerios, or sweetness using sugar water, or time using seconds between taps on a table, etc. Do not use M&Ms and make sure that (a) your lowest base level is not too small (e.g. if using weight, the lowest base should not be less than one ounce, and (b) that your range of base levels is sufficient--i.e. your highest base level must be at least 5 times as high as your lowest. Verify that you have met these conditions. Include a photo of your experimental set-up. a. What is your modality (e.g. weight, saltiness, brightness, contrast, etc.)? b. Is your highest base level at least 5x your lowest? c. If you are using weight, say how you have determined that your lowest base level is at least one ounce. 2. Select three base levels, and indicate them here (e.g. weight of 10 M&M's, 25 M&M's, 50 M&M's). If weight, include approximate weight in ounces as well as # of things used) a. Smallest Base Level 2. Select three base levels, and indicate them here (eg. weight of 10 M&M's, 25 M&M's, 50 M&M's). If weight, include approximate weight in ounces as well as # of things used) a. Smallest Base Level b. Medium Base Level c. Largest Base Level 3. For each, do a rough Method of Adjustment to get Rough discrimination thresholds (just-noticeable-difference or jnd) for each base level. For example for the M&M, the jnd might be ' M&M for a base level of 10 M&M's. You can do this without the help of a friend. You will likely need two sets of stimuli for each base level. For example, if you are using M&M's with base levels of 10, 25 and 50, you can start with two bags of 10 M&M's and then add an extra M&M to one of them until you can just barely detect that it is heavier than the other bag. If you only had to add one M&M to the second bag then your discrimination threshold (ind) for a base level of 10 M&M's is 1 M&M. Repeat this for the other two base levels. a. Rough jnd for Smallest Base Level b. Rough ind for Medium Base Level c. Rough jnd for Largest Base Level 4. Now for each base level, do a Method Constant Stimuli, selecting comparison levels at the base plus your rough jnd; at the base plus half of your rough jnd, and at the base plus twice your rough jnd. For example, for the base of 10 M&M's, If your rough jnd is 1 M&M, compare the 10 M&M's to 4. Now for each base level, do a Method Constant Stimuli, selecting comparison levels at the base plus your rough jnd; at the base plus half of your rough jnd, and at the base plus twice your rough jnd. For example, for the base of 10 M&M's, If your rough jnd is ! M&M, compare the 10 M&M's to (a) 10 % M&M's (b) 11 M&M's, and (c) 12 M&M's. Repeat each condition 10 times, in a pseudo-random fashion Record the number correct at each level. Do the same for each of the three base levels (i.e., in the M&M example, determine the approximate jnd for base levels of 10 M&M's, 25 M&M's and for 50 M&M's). This means a total of 3 base levels x 3 Comparison Levels x 10 replications each = 90 trials. You will need to have someone help you with this, so for this assignment you can work with a friend. If the friend is also a student in the class, you should each collect data on the other and turn in your own assignment. (Note that as much work as this seems, it is actually far less than most psychophysicists spendi.e. 3 base levels, 3 test points around each base level, and 10 replications per test point are each far too small for really accurate data.). Make sure your Comp Level columns are in increasing order. Make sure (a) your low level is half a rough jnd above the base, (b) your Medium level is a rough jnd above the base, and (c) that your high level is two rough jnds above the base, Base Level: Base Level: Base Level: Comp. Level Correct Comp. Level Correct Comp. Levelf Correct LOW Med High s. Plot the data as three crude Psychometric functions (one for each base level), as in Figure 1.16 or in the sample spreadsheet in the Resources under the "Assignment Copies folder. Note that in your graph, the percent on the vertical axis will typically not go far below 50, since 50% is what one would expect from random guessing. Instead of the smooth curve, simply connect the dots with a straight line. You can use the Excel spreadsheet from the Canvas Files as a model. 6. In order to increase the resolution, draw separate graphs for each of the base levels. You can use the Excel spreadsheet from the Canvas Files as a model. 7. For each base level, determine the threshold by drawing horizontal dashed lines from 75% to the graph, and then a vertical line down to the X-axis. If the lines on the graph cross 75% twice, use the second crossing (i.e. the one further to the right). If they do not cross at all, do your best to extrapolate. For example, if all percent corrects are above 75%, extrapolate down on the left side. If all percent corrects are below 75%, extrapolate up on the right side. Actually draw these lines in on the submitted spreadsheet. Read the threshold off of the axis, which should have the units of your base level plus increment (e.g. 10% M&M's 11 M&M's, and 12 M&M's. Note that if you do not reach 75% correct, you may need to extrapolate your graph to get to 75% (either upwards or downwards as necessary. If your data are super-noisy, do your best to figure the 75% point. In the example from the sample spreadsheet, you would get the following approximate thresholds: a. Base level of 10: Threshold = 11.5 b. Base level of 25: Threshold = 27.3 c. Base level of 50: Threshold = 54.5 a. Base Level of Threshold - b. Base Level of Threshold = Base Level of Threshold - 8. Convert the thresholds to jnds by subtracting off the base levels. In the example from the sample spreadsheet, you would get the following inds: a. Base level of 10: jnd = 11.5 - 10 = 1.5 b. Base level of 25: jnd = 27.3 -25 = 2.3 c. Base level of 50: jnd = 54.5 - 50 = 4.5 would get the following inds: a. Base level of 10: jnd = 11.5 - 10 = 1.5 b. Base level of 25: jnd = 27.3 -25 = 2.3 c. Base level of 50: jnd = 54.5 - 50 = 4.5 a. Base Level of jnd b. Base Level of jnd c. Base Level of jnd 9. For each base level, determine the Weber fraction=jnd baseLevel. In the example from the sample spreadsheet, you would get the following Weber fractions: a. Base level of 10: Weber fraction = 1.5/10 = 0.15 b. Base level of 25: Weber fraction = 2.3/25 = 0.092 c. Base level of 50: Weber fraction=4.5 / 50 =0.09 a. Base Level of Weber fraction b. Base Level of Weber fraction c. Base Level of Weber fraction 10. Weber's Law says that the three fractions should all be equal. How close are they? Are they at least closer to each other than the actual jnds? Do you feel that Weber's Law has been validated (recognizing that there will be variability due to not having tons of data)? Support your answer. (In the above example, you should note that while the Weber fractions are not identical, they are reasonably close. Furthermore, while the jnds show an increase with increasing base level, the Weber fraction values do not). First Assignment: Psychophysics and Weber's Pick a sensory modality for which it is fairly easy for you to precisely manipulate quantity over a wide range. Some suggestions are weight, sweetness, time and saltiness. It will be helpful to use Excel for this assignment. If you do not have Excel, you can get it for free (https://umdearborn.edu/offices/information-technology- services/software-computer-store). Note that there is a sample spreadsheet (in Canvas Files, called Sample WeberData.xls, that you can use as a starting point (but of course you must fully modify it to fit your assignment). You may not use the same base levels as in the sample spreadsheet. 1. Indicate which modality and thing you have chosen. For example you can indicate Weight using grains of rice or Weight using Cheerios, or sweetness using sugar water, or time using seconds between taps on a table, etc. Do not use M&Ms and make sure that (a) your lowest base level is not too small (e.g. if using weight, the lowest base should not be less than one ounce, and (b) that your range of base levels is sufficient--i.e. your highest base level must be at least 5 times as high as your lowest. Verify that you have met these conditions. Include a photo of your experimental set-up. a. What is your modality (e.g. weight, saltiness, brightness, contrast, etc.)? b. Is your highest base level at least 5x your lowest? c. If you are using weight, say how you have determined that your lowest base level is at least one ounce. 2. Select three base levels, and indicate them here (e.g. weight of 10 M&M's, 25 M&M's, 50 M&M's). If weight, include approximate weight in ounces as well as # of things used) a. Smallest Base Level 2. Select three base levels, and indicate them here (eg. weight of 10 M&M's, 25 M&M's, 50 M&M's). If weight, include approximate weight in ounces as well as # of things used) a. Smallest Base Level b. Medium Base Level c. Largest Base Level 3. For each, do a rough Method of Adjustment to get Rough discrimination thresholds (just-noticeable-difference or jnd) for each base level. For example for the M&M, the jnd might be ' M&M for a base level of 10 M&M's. You can do this without the help of a friend. You will likely need two sets of stimuli for each base level. For example, if you are using M&M's with base levels of 10, 25 and 50, you can start with two bags of 10 M&M's and then add an extra M&M to one of them until you can just barely detect that it is heavier than the other bag. If you only had to add one M&M to the second bag then your discrimination threshold (ind) for a base level of 10 M&M's is 1 M&M. Repeat this for the other two base levels. a. Rough jnd for Smallest Base Level b. Rough ind for Medium Base Level c. Rough jnd for Largest Base Level 4. Now for each base level, do a Method Constant Stimuli, selecting comparison levels at the base plus your rough jnd; at the base plus half of your rough jnd, and at the base plus twice your rough jnd. For example, for the base of 10 M&M's, If your rough jnd is 1 M&M, compare the 10 M&M's to 4. Now for each base level, do a Method Constant Stimuli, selecting comparison levels at the base plus your rough jnd; at the base plus half of your rough jnd, and at the base plus twice your rough jnd. For example, for the base of 10 M&M's, If your rough jnd is ! M&M, compare the 10 M&M's to (a) 10 % M&M's (b) 11 M&M's, and (c) 12 M&M's. Repeat each condition 10 times, in a pseudo-random fashion Record the number correct at each level. Do the same for each of the three base levels (i.e., in the M&M example, determine the approximate jnd for base levels of 10 M&M's, 25 M&M's and for 50 M&M's). This means a total of 3 base levels x 3 Comparison Levels x 10 replications each = 90 trials. You will need to have someone help you with this, so for this assignment you can work with a friend. If the friend is also a student in the class, you should each collect data on the other and turn in your own assignment. (Note that as much work as this seems, it is actually far less than most psychophysicists spendi.e. 3 base levels, 3 test points around each base level, and 10 replications per test point are each far too small for really accurate data.). Make sure your Comp Level columns are in increasing order. Make sure (a) your low level is half a rough jnd above the base, (b) your Medium level is a rough jnd above the base, and (c) that your high level is two rough jnds above the base, Base Level: Base Level: Base Level: Comp. Level Correct Comp. Level Correct Comp. Levelf Correct LOW Med High s. Plot the data as three crude Psychometric functions (one for each base level), as in Figure 1.16 or in the sample spreadsheet in the Resources under the "Assignment Copies folder. Note that in your graph, the percent on the vertical axis will typically not go far below 50, since 50% is what one would expect from random guessing. Instead of the smooth curve, simply connect the dots with a straight line. You can use the Excel spreadsheet from the Canvas Files as a model. 6. In order to increase the resolution, draw separate graphs for each of the base levels. You can use the Excel spreadsheet from the Canvas Files as a model. 7. For each base level, determine the threshold by drawing horizontal dashed lines from 75% to the graph, and then a vertical line down to the X-axis. If the lines on the graph cross 75% twice, use the second crossing (i.e. the one further to the right). If they do not cross at all, do your best to extrapolate. For example, if all percent corrects are above 75%, extrapolate down on the left side. If all percent corrects are below 75%, extrapolate up on the right side. Actually draw these lines in on the submitted spreadsheet. Read the threshold off of the axis, which should have the units of your base level plus increment (e.g. 10% M&M's 11 M&M's, and 12 M&M's. Note that if you do not reach 75% correct, you may need to extrapolate your graph to get to 75% (either upwards or downwards as necessary. If your data are super-noisy, do your best to figure the 75% point. In the example from the sample spreadsheet, you would get the following approximate thresholds: a. Base level of 10: Threshold = 11.5 b. Base level of 25: Threshold = 27.3 c. Base level of 50: Threshold = 54.5 a. Base Level of Threshold - b. Base Level of Threshold = Base Level of Threshold - 8. Convert the thresholds to jnds by subtracting off the base levels. In the example from the sample spreadsheet, you would get the following inds: a. Base level of 10: jnd = 11.5 - 10 = 1.5 b. Base level of 25: jnd = 27.3 -25 = 2.3 c. Base level of 50: jnd = 54.5 - 50 = 4.5 would get the following inds: a. Base level of 10: jnd = 11.5 - 10 = 1.5 b. Base level of 25: jnd = 27.3 -25 = 2.3 c. Base level of 50: jnd = 54.5 - 50 = 4.5 a. Base Level of jnd b. Base Level of jnd c. Base Level of jnd 9. For each base level, determine the Weber fraction=jnd baseLevel. In the example from the sample spreadsheet, you would get the following Weber fractions: a. Base level of 10: Weber fraction = 1.5/10 = 0.15 b. Base level of 25: Weber fraction = 2.3/25 = 0.092 c. Base level of 50: Weber fraction=4.5 / 50 =0.09 a. Base Level of Weber fraction b. Base Level of Weber fraction c. Base Level of Weber fraction 10. Weber's Law says that the three fractions should all be equal. How close are they? Are they at least closer to each other than the actual jnds? Do you feel that Weber's Law has been validated (recognizing that there will be variability due to not having tons of data)? Support your answer. (In the above example, you should note that while the Weber fractions are not identical, they are reasonably close. Furthermore, while the jnds show an increase with increasing base level, the Weber fraction values do not)
