曝光臺(tái) 注意防騙 網(wǎng)曝天貓店富美金盛家居專(zhuān)營(yíng)店坑蒙拐騙欺詐消費(fèi)者

2. Have a 2-person team enter the data (one person reads off the data, while the other enters the data into Excel.) Each IE’s ratings should appear as a single column of numbers in Excel. It may be helpful to highlight the rows of observable behavior data in one color, and the task/skill data in another color. (This makes it easier to use the “Arrange Data” macro in the next step).

3. Use the “Arrange Data” macro to separate the 3-point observable behavior data from the 4-point task/skill data. (All macros are found under the “macros” command in the “Tools” menu). Follow the on screen prompts and the end result will be two data worksheets, one with observable behavior data (Sheet 1), and the other with task/skill data (Sheet 2).

4. Save the data file onto the C: drive for faster access while running the macros.
5. Save a copy of the data onto floppy disk (A: or B: drive) for transfer to the other computer. Once the file is loaded on the other computer, be sure to save it to the C: drive for faster operation.
6. Divide up the labor. Have one person analyze the 3-point observable behavior data, while the other person analyzes the 4-point task/skill data. Each person should make sure that he/she has only the assigned type of data (i.e. there is no task/skill data mixed in with the observable behavior data).
7. Make a visual inspection of the data. Does missing data cluster by IE or by question? If 30% or more data is missing for a given IE, delete that IE. If 30% or more data is missing for a given question, delete that question. If less than 30% of the data are missing, replace the missing data. You can use the modal (most frequently occurring) value for a given question, or you can use the “Clean Data” macro to substitute the mean value.
Agreement Analyses
1. Run the “Agreement” macro. Follow the on-screen prompts. You will be asked to enter a benchmark value (GMU recommends using .80). Print the block of all ratings and agreement indexes, sized to 1 page. (This can be done by selecting “print 1 page high by 1 page wide” from the “page setup” command in the “File” menu.)

2. Next, examine each of the individual graphs for items that exhibit low agreement. If time permits, it may be helpful to format all of the graphs so that they are on a common metric. To do this, first note which graph has the largest maximum value on the vertical axis. For each graph, change the vertical axis so that they are all on this scale. This can be semi-automated by changing the format of the first graph and then using the F4 key [repeat function] to change the format on the remaining graphs.
3. Individually select and print each graph on a separate page.
4. When all the graphs are printed, hand-label each graph with the event set, item name, and the crew name.
Systematic Differences Analyses
1. From the “Tools” menu, choose “Data Analysis”. Select “2 Factor Analysis of Variance Without Replication”. Block out the raw data, starting with the @ symbol. Click the “Labels” box. Click the “Output” button, click on the output address box, and click on the desired location on the worksheet. Click “OK”.
　
中國(guó)航空網(wǎng) www.k6050.com
航空翻譯 www.aviation.cn
本文鏈接地址：Developing Advanced Crew Resource Management (ACRM) Training: A Training Manual(105)