Significance Test

this page describes how to perform significance testing on your data.

How we calculate significance difference

We use a calculation called z-test . You can choose to perform significance testing between the Column or Row values. It's also possible to select which algorithm to use: All against all, all selected values against a control group, each value against the next value or each value against the previous value.

Bellow is the specification for the options you can use.

If you want to perform significance testing on your data you'll need to purchase the sigtesting feature.

If you want to perform significance testing against a CONTROL_GROUP you need to add unique IDs to all the columns, items and rows in order for us to identify which columns / rows you are referring to.

If request payload contains incorrect IDs in controlGroup, testGroups or it will not match with any column, item or row we will return an error message.

Example JSON

The SigOptions attribute has to be passed as part of the chart Object.

{
	"body": [{
		"slide": {
			"id": "desired-unique-id",
			"table_text": "Colors",
			"question_text": "What is your favourite color?",
			"base_text": "Base Colors",
			"layout": {
				"slide_master_index": 0,
				"slide_layout_index": 1
			}
		},
		"charts": [
			{
				"sigOptions": {
					"analysisType": "CONTROL_GROUP",
					"baseLimit": 0,
					"confidence": "90",
					"controlGroup": "unique-item-id-1",
					"status": true,
					"testBetween": "TEST_BETWEEN_COLUMNS",
					"testGroups": [
						"unique-item-id-2"
					]
				}
			}
		],
		"custom_obj": {}
	}]
}

Last updated