 |
 |
 |
|
Virtual Lab Web Lecture & Tutorial ©Copyright 2000 Tom Malloy
Internet Explorer does not allow Virtual Labs to run unless you download Sun Microsystem's JAVA plugin. For our plugin help page with a tutorial on downloading the JAVA plugin and a link to Sun's download site CLICK HERE. This is the text of the in-class lecture. You may print this text out and use it as a textbook. Or you may read it online. Read Publications about the Utah Virtual Lab
Virtual Labs is an innovative program that lets you interact with an entire simulated reality discovering its principles using scientific research methods. You will study online electronic entities called Bots. Bot brains and behavior are governed by systematic principles that you as a scientist will attempt to discover. There will be four Virtual Lab assignments. The first will require that you know only basic descriptive statistics. As you learn more sophisticated statistics, the Virtual Lab Assignments will require you to use your more sophisticated knowledge. The four assignments will require you to do three research projects (chapters) each. You will be computer graded on how well you set up your research. Navigate into the Virtual Lab from Ducks in a Row, from your virtual Desk, from the Menu, or by finding it in the King of the Cats virtual city.
You will be asked to click the button indicating which Virtual La assignment you are doing. For this lecture/tutorial we will be doing Assignment #1. Assignment #1 focuses on descriptive statistics and requires a minimal amount of statistical knowledge. Virtual Lab Assignment 1. Virtual Lab assignment 1 just requires you to use basic descriptive statistics (mean, r, regression line, interaction means). Later, when you know more statistics there will be other Virtual Lab assignments requiring you to use other statistics. For Virtual Lab Assignment #1, please read Bot Res Vol 4 "Explorations." Other Volumes required more sophisticated statistics.
Click on the Front door to enter the lab.
Move your mouse around the Library. Different course titles and instructors will appear. Click on your course. For Psychology 3000 classes, go up the stairs and click on the books on the left side of the upper level of the library.
Books will appear. Click on one of them. Its full title will appear below the graphic. If that's the book that has been assigned or that you're interested in reading, click on Next. For Assigment #1, Choose Bot Res Vol 4 "Explorations."
The following text is taken from the book titled: Bot Res Vol 4 Explorations of Variables. Independent
Variables. Language Active versus Passive Processing Finger
Skills Music Learning Styles Aromas Foot Massage Literature Overview. Read the following overview of research hypotheses. Then go on to do research investigating these hypotheses as they might apply to Bots using StatCenter's Virtual Lab. 1. Small finger muscle exercises (stringing beads, knitting, working clay) stimulate brain growth. 2. Drinking more than 8 cups (8 ounces each ) of water per day increases brain function by purifying blood going to the brain. Fruit juice and carbonated beverages won't help. The brain uses more fluids than any other organ of the body. 3. The scent of peppermint increases "positive brain waves" involved in learning. 4. Testing different ideas (exploration)--especially if many of the ideas prove wrong-- builds the brain more than does knowing the right answer. 5. Foot massage (especially between the toes) nearly triples the blood flow to the brain. 6. Classical music (in contrast to white noise and pop music) produces brain development. 7. Learning multiple languages produces brain development. 8. Hearing music and language changes the structure of the brain. 9. Parts of the brain that are not used die. BOTS "Bot" is short for "robot," and is used in cyberspace to refer to electronic entities. We have simulated a cyberspace (called "StatLand") in which some of the hypotheses (about humans in the literature overview) apply to Bots and some of them don't. It's up to you to use do research and find out which apply in StatLand. When you are done reading the literature overview in your book click "Next."
Select one chapter to read by clicking on its icon. The full title of the chapter you selected will appear in the large aqua-colored bar below chapter icons. Each chapter will pose you with a story problem (scientific puzzle) to solve by doing research. Click on the large aqua-colored bar to read the selected chapter. [NOTE: You must read three chapters and do a research project on each to complete the Virtual Lab Assignments.]
As a running example for this tutorial, we have chosen "The effect of language on development" in Bot Research Vol 4. Here is the scientific puzzle in that chapter. "Ann, a Bot researcher, is interested in studying Bot brain development and function. You are her assistant. She hypothesizes that the number of languages known increases brain development in Bots. She wants you to do research to determine if Bots who have learned more languages have more developed brains. She would like you to randomly assign Bots to different groups who learn different numbers of languages. Then she would like you to measure Bot brain development in each of the groups. Report the mean development for each group. Go to the Virtual Lab and do the research project she described. Fill out the Virtual Lab sheet, most notably the following: A) State the scientific hypothesis. B) Choose a type of study to do. C) Name the IV and DV involved in your study. D) Summarize the data pattern resulting from your research. E) Does the data pattern fit with the scientific hypothesis? F) Make an overall conclusion which you are willing to defend to your peers." The chapter gives you a scientific puzzle to solve. Read it carefully to be sure you understand what the scientific hypothesis is, what the variables (IV's, DV's, Criterion Variables, as appropriate) are, and what kind of research you want to do. You might jot down informal notes on a scratch paper. You can always come back from the lab and reread the chapter if you need to review the details. To provide some structure for you this chapter uses Parts A through G to direct your attention to several things you will need to know when you write a research report at the end of your research. When you are done reading the chapter, click "OK" to return to the list of chapters. You may then choose to read another chapter or to go to the lab. To go to the lab, just click on "To the Lab." MATCHING Different scientific questions lead to different different research strategies. For example, in the Correlation web lecture we discussed how causal scientific hypotheses are best investigated through true experiments as opposed to correlational studies. when you read a scientific puzzle in Virtual Lab you must demonstrate your ability to choose the appropriate set of research tools. You do this by going to the lab that has the right tools.
You must now make an important choice: What kind of research should you do? This choice will be graded for correctness by the computer. There are four different labs for doing different kinds of research. At the time we are introducing Virtual Labs into the course, you only know how to use two of the four labs. We have pasted warning tape over the other two labs. Review. Be sure to review the distinction between experimental and correlational research at the end of the correlation web lecture. Information in that part of the lecture will allow you to read a Chapter and decide whether you should go to an experimental research lab or a correlational research lab. The Virtual Lab Chapter will not tell you what kind of research to do. Part of the point of Virtual Labs is that you must take a scientific puzzle and decide on your own how to research it.
Click here for a detailed example of using Two IV lab
When you first use Virtual Labs you will have to decide from reading the chapter whether to do correlational research or experimental research (one IV) or experimental research (two IV's). Based on your understanding of the Correlation and the Interaction chapters, you have to choose the correct lab to do your research in. Click on the lab you chose. (A yellow dot will appear indicating you are about to enter that lab.) Click "Next" to go to that lab Causal Example. How do you decide which lab to go to? Our running example is the "Effect of Language on Development." The title already hints at experimental research, since the word "effect" is part of cause-effect thinking. The crucial information for your decision is typically found in the statement of the scientific hypothesis. In our example the crucial sentence is "She hypothesizes that the number of languages known increases brain development in Bots." The way her hypothesis is worded, it is apparent that she thinks that language has a causal effect on brain development. As you learned at the end of Correlation Lecture, you CANNOT infer causation from correlation--you have to do experimental research to hope to convince your peers that an IV causes changes in a DV. Causal Scientific Hypotheses require experimental research. So to evaluate causal hypotheses you have to go to an experimental lab. Other clues to look for (besides the kind of words used in the title of the chapter and the scientific hypothesis) include terms like IV. Independent Variables go with experimental research. Correlational Example. As a contrast, let's find a chapter that requires correlational research. Here is some of the text from the chapter titled Predicting Development from Puzzle Solving. "Frank, a Bot researcher, at Bot Research Labs (BRL) on the campus of Cyber State University (CSU) is interested in whether Bot brain development can be predicted from several other variables (Electron flow to Bot brains, Bot problem solving performance, and Bot brain state). You are his assistant. He would like you to describe any linear relationship which predicts Bot brain development from Problem Solving. He also wants you to indicate if you think there is a positive, negative or no relationship between Bot brain development and the predictor variable. In this early exploratory research he is not concerned about causation; he just wants to determine which variables are related to Bot brain development and the direction of the relationship." Notice that Frank is really interested in finding variables that are RELATED to each other. A correlational study is a fine way to discover relationship. It is also a fine way of finding a regression line for predicting one variable from another. So for this Chapter you would go to the Correlational Research lab. Back to running example. But our running example is the "Effect of Language on Development." If you study that chapter, you should conclude that the hypothesis is causal, that the IV is "Number of Languages" and that the DV is "Brain Development." So we will go to the Experimental Lab (one IV). On the graphic (way back, up above), notice that we have clicked on "One IV lab Experimental data" and that a yellow dot has appeared outside the lab door on the Virtual Lab floor plan. Click "Next."
On the lab table are the research tools available in this lab. Intervener 1000. This tool allows you to choose which of four IV's to investigate. For each IV you can choose to investigate up to six levels (groups or conditions). DV Selector. This tool allows you to choose which of four DV's to measure. It also allows you to set how many Bots that you measure from n = 2 to n = 20. Clipboard. The clip board has two functions. 1) It represents your experimental design graphically. It indicates the IV and DV you have selected. It shows the Bots all lined up in the groups you have created, ready to be measured. 2) When you measure the Bots, the clipboard also shows you the data. Click on the "Intervener 1000" Tool. Intervener 1000: Research Design
IV: # of Languages. The graphic on the right shows that the "# Languages" button has been pressed. (The top button has turned from grey to white and the "# Languages" text has changed from black to blue). Also, six switches have appeared. Each switch controls one level of the IV. In this example, the levels of the IV are the various numbers of languages you have taught the Bots. The number of languages ranges from 1 to 6. Play with flipping the switches up (on) and down (off). In the graphic three switches are turned on (flipped up) and three switches are left off (flipped down). This means you're planning to have three groups of Bots. One group will know only one language, another will know three languages, and the third group will know five languages. Clipboard. At the bottom of the clipboard are three buttons. Press the one that says "Res Design." When you do that the clipboard will show you the three groups of Bots you've created. Above each group is the number of languages that group will learn (one, three, or five). Scientific Puzzle. The "Effects of Language on Development" chapter said, "She wants you to do research to determine if Bots who have learned more languages have more developed brains. She would like you to randomly assign Bots to different groups who learn different numbers of languages. Then she would like you to measure Bot brain development in each of the groups." So notice that the IV is # of Languages." Also notice that she said nothing specific about how many groups to create, or how many languages to use. So you are free to use your judgment about how many levels of language (groups) to investigate. Review. First you choose which lab to go to, based on the type of research required. In this example, assuming you have chosen to do an experimental study, you must choose the IV and its levels correctly. Some chapters will be very specific about which levels of the IV to use. Others, like the present example, leave the levels up to you. You will have to use your understanding of this class and apply that understanding to the scientific puzzle proposed in the chapter you have selected. Each experimental chapter requires you to research a particular IV and its levels. For later. The Intervener 1000 will also allow you to select whether you be running a study with independent means or correlated means (switch at bottom left of blue panel). But this switch should be left to independent means until later in the class when you learn about the t-tests and F-tests that require the use of dependent or correlated means.
The graphic on the left shows that we have selected all the levels from one to six languages. Each level indicates the number of languages known by that group of Bots. The experimental design on the graphic has changed to reflect this. As we continue the running example, we will use a research design with six levels of language.
The graphic on the left shows a case in which the wrong IV has been chosen for the Chapter "Effects of Language on Development." Music has been chosen as the IV instead of Number of Languages. You can see that six levels of Music have been selected. Music is an interesting variable, and the levels of music look intriguing, but Music was not the IV asked for in this chapter. Grading. The computer knows which chapter you have selected and it knows which IV is correct for that chapter. So it will grade you on your choice of IV. Click on Close at the bottom of the (blue) Intervener 1000 panel (left). Click on the DV tool on the desk.
DV Tool. You'll want to measure some aspect of the Bots in your study. The DV tool allows you to select which of several dependent variables to use when you measure the Bots. The DV Tool also allows you to select how many Bots are in each group of your study. If you have independent groups, then each group can have a different number of Bots. [For later: If you have correlated measurements, then the number of measurements in group must be the same in each column of numbers.] DV: Brain Development. The current graphic shows that we have selected "Brain Development" (shown in blue text at the top of the DV tool) as our DV. You choose your DV by clicking on the knob. Clicking on the knob turns the knob. Number of participants per group (n). After you select your DV, you can select the number of Bots who participate in each group of your research design. Notice on the graphic that we have chosen n = 15 for the group of Bots who learn three languages. All other groups have been left to n = 10 which is the default n. Each time you change n in any group you have to push the "Res Design" button to see the Bots again. There is no correct number of Bots for any puzzle. The data tend to be less stable with fewer Bots and more stable with more Bots.
The current graphic (left) shows a case where Brain State has been chosen as the DV. (It also shows that Music has been chosen as the IV.) Brain State is only "wrong" in the sense that it was not the DV asked for in the "Effects of Language on Development" chapter. Grading. The computer keeps track of which chapter you selected and will grade your choice of DV.
When you have chosen the correct research design for the "Effects of Language on Development" chapter, click on the "Measure Bots" button.
IV & DV. Notice on the (white) clipboard at the top the IV is listed as "# Languages." Notice also that the DV is listed as Brain Development. These are the IV and DV which are asked for in the "Effects of Language on Development" chapter. Levels of IV (Groups). Notice that we have set up six groups. At the top of each group is the number of languages that that group learned. Data. All of the Bots which were displayed in the research design have now been measured; there is a number for each of them. Notice that for the group that has learned three languages there are more data points. That, of course, is because we put more Bots in that group. Mean and Standard Deviation. At the bottom, below each group, is the mean and standard deviation for all the groups. Pop up graph. A small window pops up showing a graph of the means for each group in the research design. The height of each colored bar corresponds to the mean of each group. The small thin black line (which looks like a T) at the top of the colored bar shows the size of the standard deviation for that group. Simulated Data. You will get different data every time you press the "Measure Bots" button. Your data will also be different than any other students. And your data will be different from that shown on the graphic above. This is like real research in that the data come out a little different each time the research is replicated. The data are simulated by sampling from Normal Probability Distributions like we did with the Double Sample Tool or in the Detect Difference Game. In those tools, there were only two normal distributions. In Virtual Lab there is one normal distribution for each group you create. So the number of distributions is equal to the number of groups you created. That is, the data for each group comes from a different normal distribution. Remember how Double Sample looked; there were two probability distributions. Now there are as many normal distributions as there are groups. Virtual Lab is like the Detect Difference Game in that you don't see the normal populations, just the data. It's different than the Detect Difference Game in that you will never see the populations. You are like a scientist; you see only the data and you have to decide what the results mean.
The graphic to the right shows results for a design where the IV is Music and the DV is Brain State. It's not really "wrong." It's just not the design and results that were requested by the "Effects of Language on Development" chapter. The point is, Virtual Labs lets you create a very large number of research designs. You are free to do whatever research you want. But for the purposes of grading, the computer has a very specific research project in mind for each chapter in each book. We suggest you play around and explore doing all kinds of research. Then when you are ready, conduct a research project for a grade. Example for the Correlational Research Lab Bot Res Vol 4, "Predicting Development from Prob Solving"
Some chapters, of course, simply ask you to find relationships among variables or to predict one variable from another. In either case you will want to choose the Correlational Lab on the Virtual Lab floor plan. Notice that on the graphic the "You are Here" yellow dot is now about to enter the Correlational Lab for Measurement Data. Click "Next." If you choose the correlational lab, then things will work a little differently than they did in the experimental lab. We will just briefly go over some of the obvious differences here. But we will not go into a great deal of detail because most everything works the same in both labs. For the example below, in Bot Research Vol 4 I chose the "Predicting Development from Puzzle Solving" chapter. This chapter asks you to find a regression equation for predicting Brain Development from Problem Solving ability. Text of Chapter. "Frank, a Bot researcher, at Bot Research Labs (BRL) on the campus of Cyber State University (CSU) is interested in whether Bot brain development can be predicted from several other variables (Electron flow to Bot brains, Bot problem solving performance, and Bot brain state). You are his assistant. He would like you to describe any linear relationship which predicts Bot brain development from Problem Solving. He also wants you to indicate if you think there is a positive, negative or no relationship between Bot brain development and the predictor variable. In this early exploratory research he is not concerned about causation; he just wants to determine which variables are related to Bot brain development and the direction of the relationship. Nor is he concerned about issues of statistical conclusion validity (significance) at this stage." Notice that this chapter asks you to find a regression equation with Problem Solving Ability as a predictor (X) and Brain Development as a DV (Y).
The Correlator 500 Tool allows you to select a predictor variable (top knob) and a dependent variable (bottom knob). It also allows to select the number of Bots. The Bots will each be measured twice. Regression Studies. If are doing a regression study, use the top knob as your predictor variable (X) and then use the data to calculate a regression line. Simple Correlational Studies. If you are just finding a correlation between two variables, don't worry about the label "predictor" on the top knob. Just use the two knobs to select your two variables. Then use the data to calculate a correlation coefficient. The results in a correlational study look a little different than those in an experimental study. There are only two columns of numbers because you have just one group of Bots, each measured twice. (Review the correlation web lecture for a definition of a simple correlational study.) Each column of numbers has a Mean (M) and Standard Deviation (S). Pop up window. The pop up window shows a scatterplot of the two DV's (or one Predictor and one Criterion). Further Data Analysis. If you want to calculate r or the regression equation, you can (1) do that by hand which is a bother. Or (2) you can use "StatTool Tool". StatTool
Let us continue with the Regression Research Example (Predicting Development from Problem Solving). To get started press "Close" at the bottom of the (white) clipboard. You will something like the current graphic. Below the table, on the bottom and in the center, is a grey button labeled "StatTool." It is circled in red in the graphic. Press it. Your data are automatically imported into StatTool which will do any number of statistical analyses upon request.
If you scroll down this front page, you will find a "cheat-sheet," quick summary of instructions in case you are using StatTool when your not in this lecture. The quick summary is printable. Press "Analyze Data."
Two windows pop up. The Raw Data window is pale yellow and has the exact same data in it that you measured on your Bots. [The other window is white and has nothing in it. It is the Statistical Results window and we will talk about it below.] Menus. Along the top of the Raw Data window are a series of drop down menus. [The "File" menu is mostly used to import data from the Practice Homeworks so you can have all the answers to those problems. The other menus are for statistical analysis and graphing. The "Des Stat" menu allows you to get measures of Central Tendency and Variability. The "Cor Stat" menu allows you calculate r and the regression equation. The other menus will become relevant later in this course, but they allow you to calculate various t, chi-square, and F tests. There is also a graphing menu. For our example, the graphic above shows that the Cor Stat menu has been selected and that it has two choices--correlation and regression. The regression choice has been highlighted in dark blue. Click "Regression" on the Cor Stat menu.
The current graphic shows a close up of a part of the Statistical Results window and the Raw Data window. You can see that the predictor (X) is Puzzle Solving Ability and the DV (Y) is Brain Development. The prediction equation is Y'=0.7668X-14.2839. If you want r, then go back to the Cor Stat menu and select "Correlation." Immediately the result for r will appear in the Statistical Results window. Using StatTool is a lot easier option than hand calculation for getting complicated statistical results. You can use it for all types of research designs. Back to "Effect of Language on Development" Now let's return to our running example "Effect of language on development. We had selected our IV to be Number of Languages and our DV to be Brain Development. We had created a research design, selected n for each group and measured the Bots.
We've got our results. What do we do next? In this case we don't need to use StatTool because we are only doing exploratory research and all we need are the means and standard deviations, which are available on the clipboard. We also have a nice graph of the data pattern in the pop up window. Now we have to write a research report and make a set of scientific conclusions. Click "Close" on the bottom left of the (white) clipboard. You can also close the DV Tool if you want. This will return your full view of the Experimental Lab.
At the bottom right of the Experimental Lab is a grey button labeled "Conclusion." Click it.
Conclusion Applet
Basic Information. On the upper left of the form (in yellow box on the graphic) will be your login name (e.g., Jake), and the virtual lab assignment number you are working on. The database enters all this automatically, so you don't have to. 1) Select your course section and/or lab. In the blue box on the graphic, you can see that there is a drop down menu that lets you select your Lab TA or instructor. Please make this selection so that your record shows the correct instructor or TA. The Lantern starts out next to this drop-down menu. When you complete step 1, the Lantern wii move down to step 2. 2) Which
Chapter and Puzzle? Notice that the form has three check boxes, one for each of the three chapters/puzzles you are required to do. The illustration(in red box) shows the form with the top box checked but before any conclusions have been submited. Check any of the boxes (it doesn't matter which one). [ WARNING:
You will not see the score for the chapter you are working on when
you check the box next to the first, second or third chapter. The
score will not appear until you fill out this form completely and
have pressed the "Submit Score and Get Printable Worksheet"
button. Your work for the research project you are working on is
not submitted until you press this button.] Let's look at how theform will look after an exmple student, Jake, has already submitted all three chapter/puzzles for Assignment 1.
In the exmple ilustrated here, the database shows the score Jake already has earned for the first, second and third chapters he chose. Notice that Jake got only 75% on the chapter at the top of the list (Effect of Language on Development). If a score is lower than you like you can submit another one right over the top of it. You can do the same chapter/puzzle over again, OR you can do a different chapter. So you can keep doing research until you get all three scores as high as you like. NO, NO. The only thing you are not allowed to do is submit the same chapter more than once on the list. You cannot use the same project to count for more than one grade. For example, Jake could submit the Effect of Language on Development again to see if he can get 100%. Or he can choose a dfferent chapter and submit it in the top place in the list right over the Effect of Language on Development. Now click next at the bottom right corner of the page. The Next button will not work unless you have done steps 1) and 2) above. The Next button is enabled when the little lantern moves down next to it (as it is in the illustration).
Type
the scientific hypothesis. On the illustration, circled in blue,
is where you should type in the scientific hypothesis. Remember,
the scientific hypothesis is important in deciding what kind of
research is appropriate. Chose type of Research. On the grafic, circled in red, is a pull-down menu that allows you to select to select the type of research you just did. The choices are Correlation, Regression, Experimental one IV, Experimental two IV's, and Frequency data. Select the Appropriate Variables. In the green and purple boxes in the illustration are the locations of pull-down menus are that allow you to select the variables you just investigated. Select Statistics. In the black box is where you choose from a pull-down menu the kind of statistic which is appropriate for your research. This is an important learning feature. One of the most frequently asked questions asked me over 30 years of teaching statistics has been about how to learn when to use which statistic. Not how to calculate them, but how to learn to use them in the right way. Virtual Lab gives you a completely open ended context for practicing this important skill. So choose from the drop down menu which statistical analysis you did. For Virtual Lab Assignment 1, you only have to choose between the mean, correlation, regression interaction means. Grading. Your grade will be based on choosing the correct type of research, the correct variables, and the correct statistical analyses. Submit Conclusion. In the teal box is the button which allows you to submit your conclusions after you have filled in all the other boxes. Printable Research Report When you submit a conclusion you automatically get a printable HTML page that acts as a guide line for a reseach report. You can print the HTML research report by clicking "Print" at the top of your web browser. This research report includes the text of the the chapter, as well as the answers you entered (which have been computer graded). It also includes spaces to summarize the data and make scientific conclusions. Print and completely fill out these research reports. As you continue through the course and do several Virtual Lab assignments, you will find these research records to be invaluable study aids. Some instructors may require you to turn them in or discuss them with other students. [Summer 2000: You will not be required to turn in the research reports.] Keep a copy of all your research reports.
Researcher Identification. The person who carried out the research project is identified in the top left corner of the research report. This includes your name (in the graphic, "Bilbo") and your course number and section. Project Identification. You will also see the title of the book you chose, the name of the chapter you chose and the text of the chapter. This information clearly identifies the name and nature of the research project. Scientific Hypothesis. Below the chapter text you will see your version of the scientific hypothesis. Computer Graded Answers. Below your hypothesis, you will find an output of your answers, including the study type, the variables, and the statistics you used.
The middle section provides spaces for you to record your basic findings. Record the Descriptive Statistics. It is crucial for any scientist to have a clear record of her/his findings. You can later refer to these findings as you converse with other scientists and think about other studies that you do. This section should include whatever summary statistics are appropriate to your study (M, S, r, regression line, etc.)
Be sure to label the vertical and horizontal axes on your graph (the pop up window does not do that). The vertical axis is Brain Development. The horizontal axis is number of languages starting at one on the left and going up to six on the right. Does the data pattern fit the scientific hypothesis? Now that you have the results of the study clearly in mind, ask yourself a simple, common sense question. How do the scientific hypothesis and the data fit with each other? For example, does the scientific hypothesis predict the data? Does the data contradict the scientific hypothesis? Do the data and the scientific hypothesis appear unrelated? In the running example, the scientific hypothesis is something like "Learning more languages increases brain development in Bots." Look at the graph above. It is apparent that as you move from one language (left side of the horizontal axis) to six languages, brain development increases. This is what the hypothesis predicted. Statistical Significance. There is a horizontal line across the research report. The questions below this line are for later in the course. The questions have to do with what is called "statistical significance." When you do the first Virtual Lab assignment, you don't have to worry about filling in the answers to these questions below the line. Later, when you learn about t-tests, chi-square tests and F-tests, these questions will be important.
Summarize your Scientific Conclusion. Whether you skip the questions about statistical significance or are advanced enough to answer them, you should jot down an overall scientific conclusion that you think can be made from your research. In our example, we might conclude that learning languages increases the brain development of Bots. Percent Correct. At very bottom of the research report, you get feedback about your grade. There are four computer graded components: Type of Research, two variables, and statistical analysis. You can earn 25% for each. No Explicit Feedback. Other than your percent correct, you are not given feedback about what the type of research, variables, and statistics were "correct." That is because we want this to be a discovery process. Scientists aren't told what the "right" answer is. They just do their research and make their conclusions. They eventually get the opinions of other scientists, but there is no computer to say what the correct answer is. We, of course, are working in a "simulated" reality, so there is a "correct" answer (in the opinion of your instructor, anyway). You can go back to the lab, and re-do your research, sharpening your ideas and your ability to design the research to correspond to what was asked for in the chapter. You can do your research as many times as it takes to get 100%. That is, you can discover the principles of doing good research. Depending on the nature of the class you are taking (classroom, online) and your instructor preferences, you may or may not be asked to hand in your completed research report. Now go back, read another chapter, and solve another puzzle. You must complete three research assignments for each Virtual Lab Assignment. Two IV Lab Study Interactions between two Independent Variables Example: Music and Learning Style Select the book titled Two IV's. In that book select the chapter (or puzzle) called "Music and Learning Style." Ann, a Bot researcher, is interested in studying the flow of electrons to Bot Brains. LEARNING STYLE. Based on exploratory pilot results she hypothesizes that a Bot's Learning Style will permanently affect the flow of electrons to Bot brains. She thinks in particular that always memorizing answers will decrease electron flow to the brain. She thinks that using complex formulas will slightly increase electron flow. Engaging in Trial and Error learning will increase electron flow even more. And, finally, she thinks that engaging is Discovery processes (like Virtual Lab) will increase electron flow even more than trial and error. MUSIC. She also hypothesizes that how Bots process music over a long period of time will affect the flow of electrons to their brains. Simply Listening to music won't affect flow one way or the other. But actively processing music by dancing and playing an instrument will both increase electron flow. She is not sure which, dancing or playing, will increase flow more. MUSIC by LEARNING STYLE INTERACTION. More important, she hypothesizes that Music and Learning Style INTERACT to affect electron flow to Bot brains. She thinks that the the interaction is complex and wants you to research only a small part of it. She thinks that the effect of IV1, Learning Style (Memorizer versus Discoverer), on the DV, electron flow, will depend on IV2, Level of Musical Skill. If a Bot only listens to music then a Discovery Learning style will reduce the electron flow to the brain. But if a Bot plays music then a discovery learning style greatly increases electron flow to the brain. IN SHORT: If Bots Play Music, then electron flow should be higher for Discoverers than for Memorizers. But if Bots only Listen to Music then electron flow should be higher for Memorizers than for Discoverers. She wants you to do research to determine if the interaction she has described shows up in the data. Specifically she wants you to run a study in which Bots either Play music or Listen to music and have either a Memorize or a Discovery learning style. This should
give you a study with four groups. She predicts an interaction in which those Bots who Play have lower flow than Listeners if they are Memorizers but those Bots who Play have higher flow than Listeners if they are Discovers. Go to the Virtual Lab and do the research project she described. Collect date, make up a table of means, and make a graph of the results. She only needs the group means that show the interactions. Then fill out a research conclusion report. (Just choose "Interaction Means (2 IV's)" as your data analysis).
After you have read and digested the research puzzle, GO TO THE TWO IV LAB. The Laboratory
Tools 3 Tools Notice that as in the One IV lab, there are 3 tools, a tool to select the DV, a tool to select IV's and a clip board to show you the design and data. In this case, the IV tool allows you to simultaneously select 2 independent variables.
The IV Tool and the Research Design Clipboard IV's: The figure below shows on the left the IV tool open. Notice that "Musical Skills" (IV1) is selected as one of the two IV's. Levels of Musical Skills that have been selected are "Listen" and "Play." The other IV that has been selected is "Learning Style" (IV2). The levels of Learning Style that have been selected are "Memorize" and "Discover." These selections fit with the research puzzle described above. Clipboard: On the right of the figure below is the clipboard showing the research design. For example, across the top is shown IV2 with its levels and down the side is IV2 with its selected levels. The four blue squares represent the four groups formed by combining the two levels of IV1 and two levels of IV2. Notice that each group (blue square) has the mean of that group printed on it. If you click on a blue square you will see the raw scores that produced that group mean.
StatTool
Table of Means and 2 way ANOVA Summary Table
If you click on "Analyze Data" your data will imported into StatTool where you can analyze it. The figure shows a table of means generated for the current research project. It also shows the results of a 2-way ANOVA.
StatTool
Graphs StatTool also lets you graph the results of your experiment. The figure shows the data for the current example graphed with Learning Style on the horizontal axis and Musical Skills as two lines on the graph. You can
choose to put either IV on the Horizontal Axis This graphic shows that you can put Musical Skills on the horizontal axis and put Learning Style as two lines in the graph. Which way you choose to graph the data depends on what point you are trying to comunicate and what looks best to you. Now go back, read another chapter, and solve another puzzle. You must complete three research assignments for each Virtual Lab Assignment.
|
|
Which
Assignment?
Library
Books
Example:
Bot Research VOLUME 4: Explorations of Variables
Chapters
Example: The Effect
of Language on Development
Virtual
Labs Floor Plan
Experimental
Research One IV. One of the labs allowsyou to investigate the
effect of one IV on a DV. At the time we introduce Virtual
Labs, this is one of the labs you should be able to use.
Experimental
Research Two IV's. A second lab allows you to investigate the
effect of two IV's on a DV. You have learned about this in
the interaction lecture. You will learn more about this kind of
research toward the end of the course when we learn about a topic
called "Two way analysis of variance." 
Correlational
Research. Another lab is for doing correlational studies with
measurement data Like height, weight, IQ, etc.). 
Categorical
Variables and Frequency Data. The last lab is for doing research
with categorical variables and frequency data. You will learn about
this kind of research when we study a topic known as "Chi-square."
Experimental
Lab for ONE IV 
This
tool allows you to choose an independent variable (by pressing the
buttons) as well the levels of the IV (by flipping the switches).
Choosing an IV
and its levels creates several groups of Bots, each treated with
one of the levels of the IV.
Six
levels of Language
Counter
Example: Wrong IV 
DV
& Research Design 
Counter
Example: Wrong DV
Results
on Clipboard
Counter
Example: Wrong Results
Correlational
Lab. In correlational
labs there are only 2 tools, the Correlator 500 and a Clipboard.
They work in a similar way to the tools in the experimental labs.
The Correlator 500 allows you to select two DV's or, if you choose
to do a regression analysis, a predictor variable and criterion
(dependent) variable.
Selecting
DV's and getting Results. 
Instructions
for using StatTool are not part of this lecture. But we will
briefly go over the main steps here because it is simple to use.
StatTool's
front page will pop up. (StatTool is also sometimes called Analysis
Tool.) StatTool is part of a set of online resources called StatCenter.
Raw
Data Window
Statistical
Results
Fill
in the Page 1 of the Conclusion Form. A
simple program (applet) will pop up giving you a form to fill out.
What you enter on this form will go to the database and will affect
your grade. A little lantern symbol will appear next to the part
of the form you should be fillig out. When you compete one part
of the form, the lantern jumps down to the next part of the form.
Step
2 AGAIN: Which Chapter/Puzzle? For each Virtual
Lab assignment you are required to do a research project based on
three DIFFERENT chapters. You are free to choose among
many chapters for research projects; but you must research the puzzles
in three different ones. 
Page 2 of the Conclusion Form. The Lantern will guide you through
steps on the final page of the Conclusion Form.
Top
section of research report
Middle
section of research report
Sketch
a graph of the results. It is invaluable to think about your
data visually as well as symbolically. Draw a little informal scatterplot,
regression line, or graph of the group means in the space provided.
In our running example, "Effect of Language on Development,"
a sketch of the group means similar to the one that appeared in
the pop up window when you measured the Bots would suffice. This
is not meant to be busy work, so don't worry about draftsmanship.
Just sketch a graph which allows you to "see" the data
pattern. 
Bottom
section of research report
