Project 3 Attention General Psychology Lab 8 Project 3 Attention (Print lab and How to calculate and Report a t-test, bring both to class) The process of attention is a part of our everyday waking consciousness. Attention is often defined as a concentration of mental activity to process incoming sensations. The amount of attention it takes to process information varies with the difficulty of the task, our experience with the task, and the number of sensations we are trying to process. Research has found that we can pay attention to several sensations at one time if we can process some of the incoming information automatically (, 1992). Studies in psychology show that there may be several types of attention; two of these are: divided attention and selective attention. Divided attention means we can distribute our attention to perform more than one task at a time. However, if we try to do too many tasks at once, we won’t do any of the tasks well. Selective attention means trying to pay attention to one stimulus while ignoring the other stimuli we are receiving. Have you gone to the circus and tried to follow all three rings at once? How often do you try to study with the TV on and find that you are paying more attention to the TV than your studies? Does paying attention to one object or thing reduce the amount of attention you have available to do something else? If we have a limited attention capacity, how is it that many of us can easily do more than one thing at a time? Automatic processing allows us to carry out more than one activity at a time. Automatic Processing Think back to when you first learned to drive a car, do you remember how much you had to think about? You had to stay in the center of the lane. You had to check the speedometer all of the time. You had to concentrate to plan your turns or parallel park. Now think about how you act when driving today. You probably are not even aware of these things while behind the wheel. This is an example of automaticity the apparently effortless, involuntary processing conducted without intention (Zimbardo & Weber, 1995). Automaticity occurs when one or more of the following happen: task performance requires less effort task can be done with little conscious attention more than one task can be performed at the same time without interference Automaticity is generally very helpful. It frees us to do more complicated activities than we could do if we had to think about every little step involved in a task. A major problem with the phenomenon of automaticity is once you have learned to do something it is very hard to over-ride the automatic processing of information. A major reason this concept has been a focus of research is the problem humans have in breaking habits. Habits are a type of automaticity. Psychological research in this area includes investigating how habits form and how to break habits. Introducing the t-test During the last lab you learned how to enter data and calculate descriptive statistics. Today’s lab introduces you to an inferential statistic that will assist you when drawing conclusions about the data collected. A t-test measures the difference between two means and determines the probability that the difference occurred by chance. There are two basic types of t-tests: paired and independent. You use the paired t-test when each participant contributes more than one data value to the data set. The independent t-test is used when each participant contributes one value for each variable in the data set. Example of how a t-test works. A researcher thought that the amount of sleep the participants had before taking part in a memory study would effect the number of words they could remember. There were two conditions in the study, the same participants were used in each condition. In Condition 1 the participants were allowed eight hours of sleep the night before the memory test. In Condition 2 the participants were allowed four hours of sleep the night before. The number of words remembered in each condition were: Amount of Sleep Eight Hours Four Hours Number of words recalled M = 12.00 10.00 SD = 3.25 2.94 The difference between the means was 2.00. Is this difference the result of chance (reflecting the normal variation among people) or can we say that less sleep caused the difference? When the difference is the result of chance, the researcher must conclude that there is no difference between the means; both numbers may as well have been exactly the same. How do researchers determine chance? They use the probability of what the t-test value would be if it were a chance occurrence. SPSS will help you determine this value. Researchers often want to compare the mean of the data in one condition to the mean of the data in another condition of a study. If the difference between the two means is large enough the result is not due to chance. When a researcher says ‘difference’ this does not imply the numbers were different but the numbers were statistically different. Often two means appear ‘different,’ but after using a t-test the researcher finds there is not a large enough difference to reach statistical significance. Determining whether the difference in the means of a study are chance occurrences or the result of the different study conditions is an important step in research. This determining factor is called the probability of the difference. Probability/Level of Significance Researchers in psychology do not say they have proven their hypothesis. They report the probability that the difference between the means did not occur by chance. The term level of significance reflects how sure the researcher is of the results. The probability that another study, conducted exactly like the original study, will have different results less than 5 times out of 100 is reported as p<.05 (Read: probability less than point-oh-five). In this case the researcher is 95% sure the results did not occur by chance. At times researchers use a more stringent level of significance of .01. This means that 99 times out of 100, the results will be the same. In SPSS you will find the Level of Significance as the value labeled Sig(2-tailed). The probability is related to the number of data values you have in the data set. This calculation is called the degrees of freedom (df). The computer program will calculate the df for you. The larger the df the smaller the t-value can be to reach significance. Interpreting the t-test The t-test measures the difference between two means and determines whether the difference between the groups is large enough to be significant. The df indicates the number of data values in the data set. The t-value measures the distance between the two means. The ‘p’ (Sig.(2-tailed)) value determines if the t-value is large enough to not be a chance occurrence based on what would be expected given the number of data values in the data set. Figure 3.1 The t-test analysis output t-value df p value Reporting the Results of a t-test When reporting the results of a statistical test, there are format conventions you need to follow. A t-test result is reported using a small t, followed by the degrees of freedom in parentheses, an equals sign, the t value obtained, a comma, then the probability: t(9) = 7.43, p<.05 You can read the t-test equation as: the t for nine degrees of freedom is 7.43; the probability of this result is less than five percent (or point .05). Whenever you report a t-test result you must also report the means and standard deviations for the groups you are comparing. The acceptable format for reporting the mean and standard deviations is: The group that slept eight hours recalled more words (M = 12.00, SD = 3.25) than the group that slept four hours (M = 10.00, SD = 2.94). Always report the mean and standard deviation together. Do not forget the 2 decimal reporting rule. Although the t-test will tell you if the probability of the difference is significant there are additional steps you will need to complete before you can determine if your hypothesis was supported or not. Determining Support for Your Hypothesis A hypothesis is said to be rejected or not supported, when the probability is >.05 or the means are not in the predicted direction. When the probability is <.05 and the means are in the direction predicted in the hypothesis the hypothesis is said to be supported. Hypothesis: A greater number of words will be recalled by participants when they have eight hours of sleep the night before a memory test than when the participants have four hours of sleep the night before the memory test. Study Result 1: In this study the hypothesis would be supported, the group getting eight hours of sleep recalled significantly more words than the group that got only four hours of sleep. Study Result 2: In this study the hypothesis would not be supported even though the t-tests are identical. This difference in interpretation occurs because the hypothesis stated the participants with eight hours of sleep would recall more words than the group with four hours of sleep and they did not. Study Result 3: In this study the hypothesis would not be supported because the probability of the difference was greater then .05. The difference of 2.00 occurred by chance, it was not the result of the different conditions in the study. Summary of the steps needed to determine support for your hypothesis 1. Is the ‘p’ ( Sig (2-tailed)) greater or less than .05? Yes (p < .05), check the means. No (p > .05), hypothesis is not supported. 2. Are means in direction predicted? Yes, hypothesis is supported. No, hypothesis is not supported. Project 3 Attention Lab Activity Stroop Effect Study The experiment we will do is called the Stroop Effect. Even though the principles were first introduced in 1935, this method of measuring the change in speed to complete the task is still a reliable measure of how automaticity can affect performance (VanWallendael & Matlin, 1995). When we see a group of letters our first response is to try and form them into a word. This is a fairly automatic process, but if we are asked to override this process we have to use more of our attention and this slows down our response. Stroop’s original research question in 1935 was; Why are we often unable to ignore relevant information even when we are told to do so (Stroop, 1935)? Hypothesis The time it takes to read the color of the ink in the conflicting conditioning will be significantly longer than the time needed to read the color of the ink in the same condition. The equipment needed for this study: A stop watch A set of stimulus cards Same = color of ink and color word the same Neutral = non-words printed in different colors of ink Conflicting = color of the ink is different from the color word Procedure Students will form pairs. One student will be the participant and the other the experimenter. 1. The experimenter will hand the cards to the participant in the order listed in the individual data collection form on the following page. 2. The participant will hold the card face down until the experimenter says “begin.” 3. The experimenter will say "begin" while starting the stopwatch. 4. The participant turns the card over and reads the color of the ink the items are printed with for each of the 20 items on the card. The participant should self-correct any errors, such as reading the word rather than the color. 5. When all 20 colors have been 'read', the experimenter stops the watch and records the time for that trial on the data collection form. Each pair should record the data in both lab manuals. Data Collection Forms Individual Times - Present the cards to the participant in the order below starting with Trial 1 through Trial 15. Form 1. Time to read cards in order of presentation Form 2. Summary of Individual Times by condition - use the data from Form 1 Calculate the mean time for each of the three conditions and share the data with the class. Data Analysis The analysis of the class data from today’s study will consist of descriptive statistics for each condition and a series of paired t-tests to test two hypothesis. Opening the Data File Created Last Week 1. Start SPSS from the Colvard 5095 window 2. Click on the circle in front of Open an Existing data source 3. Click on the down triangle to the right of the Look In box. 4. Navigate to the Data file on the S drive by opening the following: Server – Dvol1 on ‘Dataserv1’ (S:) Folder – Coas; Folder – PSYC; Folder - Labs; Folder - Lab Data; Folder - Your lab section Open the File Intro Lab XX. Define the Variables Define three new variables in the Variable View window - same, neutral, and conflict. Name = same, Type = numeric, Width = 8, Decimal = 3, Label = Same Name = neutral, Type = numeric, Width = 8, Decimal = 3, Label = Neutral Name = conflict, Type = numeric, Width = 8, Decimal = 3, Label = Conflict Click on the Data View tab to enter your data values. Enter the data from the Stroop study into the three new columns. The first two columns in the data set should be the alphabet data (dom and nondom) and then the three columns from the Stroop Effect study. Calculating a Paired t-test Using SPSS 1. Move to Analyze in the menu bar. 2. From the list of statistical procedures move to Compare Means. 3. From the pop-up menu select Paired Samples t-test. In the Paired Samples t-test dialog box. (Figure 3.3) a. Click once on the first variable for your first t-test. b. The name of the variable you have Figure 3.3 Paired t-test dialog box selected will appear in the paired variables box. c. Click once on the second variable needed for your first t-test. 4. When your two variables are selected, click once on the OK button to calculate the t-test. Reading the Paired t-test Analysis Window You will find the means and standard deviations in the upper portion of the window. In the sample (Figure 3.4): DOM M = 18.40 SD = 3.50 NDOM M = 8.00 SD = 1.94 Figure 3.4 Paired Samples descriptive statistics The t-test information is shown in the chart labeled Paired Samples Test (Figure 3.5). You will use the three values on the far right of this chart. Ignore the middle section labeled Paired Differences. Figure 3.5 Paired t-test output To correctly report the t-test: The t value is shown below the t. In the sample the t value is 7.429. This value is placed after the equals sign in the t-test equation. The degrees of freedom are below the df = . In the sample the DF = 9. This value is placed in the parentheses in the t-test equation. The level of significance is below the Sig (2-tailed). Compare this value to .05 and report whether it is greater than (>) or less than (<). For this sample the probability would be reported as p<.05. Report whether it is < > .05 NOT the Sig (2-tailed) value. The t-test for Figure 3.5 would be reported as: t (9) = 7.43, p<.05. Project 3 Attention Lab Report Student Name _________________________________________________________________________________ Instructor's Name ____________________________________________ Lab Section Number ________________ Class Data from the Attention Study 1. Report the descriptive statistics for the class data in each condition of the Stroop Effect study and the Alphabet study. Do not forget the 2 decimal point reporting rule (____ of 5 points) 2. Report the paired t-test between Same and Conflicting for the Stroop study. (____ of 4points) Same/Conflicting t(__________) = ______________ , p ________.05. 3. Report the t-test between the two conditions of the alphabet study. (____ of 4 points) Dom/Ndom t(__________) = ______________ , p ________.05. 4. Was the hypothesis: “The time it takes to read the color of the ink in the conflicting conditioning will be significantly longer than the time needed to read the color of the ink in the same condition” supported? (____ of 3 points) Check One Yes __________ No ___________ 5. Was the hypothesis that more letters would be written by the group using their dominant hand supported? (____ of 3 points) Check One Yes _________ No __________ 6. Name the independent variable and each of the levels of the IV in the Stroop Effect study. (____ of 3 points) IV – Level 1 Level 2 7. Name the dependent variable in the Stroop Effect Study. (____ of 3 points) 8. Which research design was used in the Stroop Effect study? (____ of 2 points) 9. Describe the steps needed to determine whether a hypothesis is supported or not supported. (____ of 3 points) If the results were: Eight hour mean (M = 12.00 SD = 3.25) t(14) = 4.53, p<.05 Four hour mean (M = 10.00 SD = 2.94) If the results were: Eight hour mean (M = 10.00 SD = 2.94) t(14) = 4.53, p<.05 Four hour mean (M = 12.00 SD = 3.25) If the results were: Eight hour mean (M = 10.00 SD = 2.94) t(14) = 2.98, p>.05 Four hour mean (M = 12.00 SD = 3.25) Trial Condition Time Trial Condition Time Trial Condition Time 1 Same 6 Conflict 11 Conflict 2 Neutral 7 Conflict 12 Neutral 3 Conflict 8 Neutral 13 Conflict 4 Neutral 9 Same 14 Same 5 Same 10 Same 15 Neutral Trial Same Trial # Neutral Trial # Conflict 1 2 3 5 4 6 9 8 7 10 12 11 14 15 13 Total Time Total Time Total Time Mean Mean Mean same neutral conflict 1 2 3 4 5 6 7 8 9 10 Mean SD Dominant Hand (DOM) Non-Dominant Hand (NDOM) Same Neutral Conflicting