Correlation - Revision history

Mtabadmin at 17:45, 7 August 2018

2018-08-07T17:45:43Z

Mtabadmin at 00:29, 4 August 2018

2018-08-04T00:29:06Z

Mtabadmin at 19:13, 1 August 2018

2018-08-01T19:13:31Z

Mtabadmin at 19:13, 1 August 2018

2018-08-01T19:13:05Z

Mtabadmin at 19:10, 1 August 2018

2018-08-01T19:10:49Z

Mtabadmin at 19:04, 1 August 2018

2018-08-01T19:04:15Z

Mtabadmin at 18:57, 1 August 2018

2018-08-01T18:57:36Z

Mtabadmin at 18:53, 1 August 2018

2018-08-01T18:53:36Z

Mtabadmin at 06:59, 1 August 2018

2018-08-01T06:59:43Z

Mtabadmin at 06:51, 1 August 2018

2018-08-01T06:51:27Z

@@ Line 29: / Line 29: @@
   '''Important: Include the question you would like to correlate the rest of the questions to at the top of the list in rows.'''
 In this example, Overall Opinion is included first, so the rest of the questions will be correlated to Overall Opinion.
@@ Line 39: / Line 40: @@
-If you receive a warning message that highlights any of your row questions, this most likely means that the question you selected does not have any stat values assigned to it. <br />To proceed with a correlation report, you’ll need to assign some type of stat values to the highlighted questions. <br />Proceed by double-clicking the highlighted question. <br />The Subset Editor will appear.
+If you receive a warning message that highlights any of your row questions, this most likely means that the question you selected does not have any stat values assigned to it. <br />To proceed with a correlation report, you’ll need to assign some type of stat values to the highlighted questions.
@@ Line 84: / Line 92: @@
 '''Heat Map - Degree of Correlation'''<br />This option assigns colors to help categorize the strength of the correlation. <br />The darker the color, the stronger the correlation. <br />The default colors of green and red identify positive correlation (green) and negative correlation (red). <br />The colors can be changed based on user preference.
@@ Line 89: / Line 110: @@
-  NOTE: The heat map colorization is not identifying significance or any other type of mathematical calculation. It is intended more as an art to categorize correlation strength.
+  '''Note: The heat map colorization is not identifying significance or any other type of mathematical calculation. It is intended more as an art to categorize correlation strength.'''
-  Green identifies a positive correlation, so the 2 questions (such as Overall Opinion and Ease of Use) change in the same direction.<br />&nbsp;<br />Red identifies a negative correlation, so the 2 questions (such as Overall Opinion and Did You have any Problems) change in the opposite direction.
+  '''Green identifies a positive correlation, so the 2 questions (such as Overall Opinion and Ease of Use) change in the same direction.<br />&nbsp;<br />Red identifies a negative correlation, so the 2 questions (such as Overall Opinion and Did You have any Problems) change in the opposite direction.'''

@@ Line 89: / Line 89: @@
-NOTE: The heat map colorization is not identifying significance or any other type of mathematical calculation. It is intended more as an art to categorize correlation strength.
+ NOTE: The heat map colorization is not identifying significance or any other type of mathematical calculation. It is intended more as an art to categorize correlation strength.
-Green identifies a positive correlation, so the 2 questions (such as Overall Opinion and Ease of Use) change in the same direction.
+ Green identifies a positive correlation, so the 2 questions (such as Overall Opinion and Ease of Use) change in the same direction.<br />&nbsp;<br />Red identifies a negative correlation, so the 2 questions (such as Overall Opinion and Did You have any Problems) change in the opposite direction.
-−
 Red identifies a negative correlation, so the 2 questions (such as Overall Opinion and Did You have any Problems) change in the opposite direction.

← Older revision		Revision as of 19:13, 1 August 2018
Line 28:		Line 28:


−	~~Important:~~ '''Include the question you would like to correlate the rest of the questions to at the top of the list in rows.'''	+	'''Important: Include the question you would like to correlate the rest of the questions to at the top of the list in rows.'''

	In this example, Overall Opinion is included first, so the rest of the questions will be correlated to Overall Opinion.		In this example, Overall Opinion is included first, so the rest of the questions will be correlated to Overall Opinion.

@@ Line 42: / Line 42: @@
-[[File:correlation5.png|800px]]
+[[File:correlation5.png|500px]]
@@ Line 48: / Line 48: @@
-[[File:correlation6.png|800px]]
+[[File:correlation6.png|500px]]
@@ Line 81: / Line 81: @@
 '''Select Confidence Interval'''<br />Options include either 95% or 99%.<br /> The default is 95.
-'''Significant'''<br />Significant values are identified in bold font. <br />A value in '''bold font''' is equivalent to other programs such as SPSS assigning an asterisk (*)to identify 95% significance and two asterisks (**) to identify 99% significance.
+'''Significant'''<br />Significant values are identified in bold font. <br />A value in '''bold font''' is equivalent to other programs such as SPSS assigning an asterisk (*) to identify 95% significance and two asterisks (**) to identify 99% significance.
 '''Heat Map - Degree of Correlation'''<br />This option assigns colors to help categorize the strength of the correlation. <br />The darker the color, the stronger the correlation. <br />The default colors of green and red identify positive correlation (green) and negative correlation (red). <br />The colors can be changed based on user preference.

@@ Line 45: / Line 45: @@
-In this example, the responses (Yes, No, etc.) do not have any numeric stat values assigned to them. Assign your own values by clicking in the column to the right of the question and enter stat values for each response.
+In this example, the responses (Yes, No, etc.) do not have any numeric stat values assigned to them. <br />Assign your own values by clicking in the column to the right of the question and enter stat values for each response.
 [[File:correlation6.png|800px]]
-If assigning a stat value is not applicable for one or more responses, you can use the subset feature to include only the responses you’ve assigned values to. In this example, plus signs [+] indicate that Yes, No, and Yes, None Checked will be included in the correlation report. Other will not be, since it does not have a stat value assigned.
+If assigning a stat value is not applicable for one or more responses, you can use the subset feature to include only the responses you’ve assigned values to. <br />In this example, plus signs [+] indicate that Yes, No, and Yes, None Checked will be included in the correlation report. <br />Other will not be, since it does not have a stat value assigned. <br />Click OK to proceed and exit out of the Subset Editor.
 Click OK to proceed and exit out of the Subset Editor.
 [[File:correlation7.png]]
 mTAB currently supports two types of correlation, Spearman’s and Pearson’s. By default, mTAB will run Spearman’s, but you can change the option, based on the descriptions for the type of data you would like to run.
 [[File:correlation8.png|800px]]
-The correlation report will run. Since a question was included in Layer (Brand question), the correlation report will include a worksheet per brand and an additional '''Report''' view sheet, which includes the correlation results for Overall Opinion for each brand.
+The correlation report will run. <br />Since a question was included in Layer (Brand question), the correlation report will include a worksheet per brand and an additional '''Report''' view sheet, which includes the correlation results for Overall Opinion for each brand.
 [[File:correlation9.png|800px]]
 Click on the Indicators icon
-[[File:correlation9-5.png|800px]]
+[[File:correlation9-5.png|500px]]
@@ Line 77: / Line 81: @@
 '''Select Confidence Interval'''<br />Options include either 95% or 99%.<br /> The default is 95.
-'''Significant'''<br />Significant values are identified in bold font. A value in bold font is equivalent to other programs such as SPSS assigning an asterisk to identify 95% significance and two asterisks to identify 99% significance.
+'''Significant'''<br />Significant values are identified in bold font. <br />A value in '''bold font''' is equivalent to other programs such as SPSS assigning an asterisk (*)to identify 95% significance and two asterisks (**) to identify 99% significance.
-'''Heat Map – Degree of Correlation'''<br />This option assigns colors to help categorize the strength of the correlation. The darker the color, the stronger the correlation. The default colors of green and red identify positive correlation (green) and negative correlation (red). The colors can be changed based on user preference.
+'''Heat Map - Degree of Correlation'''<br />This option assigns colors to help categorize the strength of the correlation. <br />The darker the color, the stronger the correlation. <br />The default colors of green and red identify positive correlation (green) and negative correlation (red). <br />The colors can be changed based on user preference.
 [[File:correlation10.png|800px]]
 NOTE: The heat map colorization is not identifying significance or any other type of mathematical calculation. It is intended more as an art to categorize correlation strength.
@@ Line 89: / Line 94: @@
 Red identifies a negative correlation, so the 2 questions (such as Overall Opinion and Did You have any Problems) change in the opposite direction.
 [[File:correlation11.png|800px]]
 Click the Export icon to save a fully editable file of the correlation report in either ''XLSX'' or other formats.

← Older revision		Revision as of 19:04, 1 August 2018
Line 39:		Line 39:


−	If you receive a warning message that highlights any of your row questions, this most likely means that the question~~(s)~~ you selected does not have any stat values assigned to it. To proceed with a correlation report, you’ll need to assign some type of stat values to the highlighted questions. Proceed by double-clicking the highlighted question. The Subset Editor will appear.	+	If you receive a warning message that highlights any of your row questions, this most likely means that the question you selected does not have any stat values assigned to it. <br />To proceed with a correlation report, you’ll need to assign some type of stat values to the highlighted questions. <br />Proceed by double-clicking the highlighted question. <br />The Subset Editor will appear.
		+

	[[File:correlation5.png\|800px]]		[[File:correlation5.png\|800px]]

@@ Line 5: / Line 5: @@
 Statistical correlation is a quantitative way to measure the direction and strength of the tendency to vary together.
-  -1.00 Perfect Correlation (both variables move in opposite directions)
+  '''-1.00 Perfect Correlation (both variables move in opposite directions)'''
-.00 No Correlation
+   '''0.00 No Correlation'''
-.00 Perfect Correlation (both variables move in the same direction)
+   '''1.00 Perfect Correlation (both variables move in the same direction)'''
 Please follow these steps and guidelines to setup and run correlation in mTAB.
@@ Line 24: / Line 24: @@
 Including a Layer question is optional. This example includes a Brand question in Layer. If a Layer question is used, it will default to the Outside Worksheets orientation and display results in the worksheet format.
-In this example, Overall Opinion is included first, so the rest of the questions will be correlated to Overall Opinion.
+[[File:correlation2.png|800px]]
-[[File:correlation2.png|800px]]
+ Important: '''Include the question you would like to correlate the rest of the questions to at the top of the list in rows.'''
-Click the bottom part of the Run (lightning bolt) icon. It will expand and offer additional Run options. Click the Run Correlation icon.
+Click the bottom part of the Run (lightning bolt) icon. <br />It will expand and offer additional Run options. <br />Click the '''Run Correlation''' icon.

@@ Line 20: / Line 20: @@
 Columns are not used in correlation. If a question is included in columns, it will be ignored.
-Including Filters or a Layer question are optional. This example includes a Brand question in Layer.
+Including a Filter is optional.
   Include the question you would like to correlate the rest of the questions to at the top of the list in rows.

@@ Line 12: / Line 12: @@
+[[File:correlation1.png|800px]]
 Begin by including questions in rows. The most commonly used question types for correlation reports run in mTAB are Likert-scale or ratings questions.
@@ Line 22: / Line 25: @@
 In this example, Overall Opinion is included first, so the rest of the questions will be correlated to Overall Opinion.
-−
 Click the bottom part of the Run (lightning bolt) icon. It will expand and offer additional Run options. Click the Run Correlation icon.
-−
 If you receive a warning message that highlights any of your row questions, this most likely means that the question(s) you selected does not have any stat values assigned to it. To proceed with a correlation report, you’ll need to assign some type of stat values to the highlighted questions. Proceed by double-clicking the highlighted question. The Subset Editor will appear.
 In this example, the responses (Yes, No, etc.) do not have any numeric stat values assigned to them. Assign your own values by clicking in the column to the right of the question and enter stat values for each response.