AML Documentation

Content

<g>

- Start a graphic presentation or define <g> attributes

<xt>

- Start a cross table or define <xt> attributes

<dig>

- <dig> Create a Dig analysis


< g > - Start a graphic presentation or define <g> attributes

To top
Start a graphic presentation or define <g> attributes. Its purpose is to start a graphic presentation by <g> or to define attributes to be inherited by subsequent <g>'s

a

Defines the main graph address to be analysed and presented

address_expression 

Specify the adresses to be shown graphically

Example: The adress expression a='\4' will show a graphic presentation of question 4.

autocode

Ascending or Descending will make an address vector of all occurencies of a numeric data type (F) in the a attribute and the values will appear ascending or descending in the graph. The H address component may be present in the a attribute in order to limit the value vector.

ascending 

The quantitative question will be shown in ASCENDING order

no 

The quantitative question will NOT be autocoded.

Example: Question 2 is age. The graph <g a='\2' autocode='descending'/> will show the frequency of all age groups in the survey in a descending order. AUTOCODE IN INTERVALS? (Link: autocode.htm)

autocodeinterval

Specify the autocode clustering

Specify any positive number

Example: If autocodeinterval='2' is specified, the autocode feature will show categories for every 2nd value

b

Defines an address that adds new dimension(s) to the graph. This address is combined with the address of the 'a' attribute and the appearance of the graph is determined by the organizing rule found in the a attribute address.

address_expression 

Specify the number of adresses to be shown in the graphs.

Example: If question 5 is household income, question 1 is gender and 4 is age groups, the graph the following expression <g a='\5' b='\1 \4'/> will show the household income split down to gender and age. If you want to include also the total, you can specify ALL, e.g. b='ALL \1'

basedoc

Documents the size of the chart.

no 

No, will not show base documentation

yes 

Yes, will show base documentation. Default value is number of interviews

int 

Number of interviews

pop 

Weighted population

The base documentation will show together with the chart title.

catdoc

Documents the size of the categories in categorical questions.

no 

No, will not show category documentation

yes 

Yes, will show category documentation. Default value is number of interviews

int 

Number of interviews

pop 

Weighted population

The category documentation will show together with the categary labels.

chartonnewslide

Show chart as chart one on a new PowerPoint slide instead as a chart number two on the same slide together with previous chart.

yes 

Will start a new slide for this chart.

no 

Will continue with chart on same slide as previous chart.

This attribute will work together with the attribute: pppSlideLayout='ppLayoutTwoObjects'.

colorindex

Specify the colors in the chart.

#NNNNNN 

Define the colors to be used in the chart with hex-color specifications. Add a space between the colors.

The colors will override the existing template in e.g. a PowerPoint presentation.

deciavg

The number of decimals to be used for average and fractile values

Number of decimals. A non-negative integer of maximum 9 digits.

decipct

The number of decimals to be used for percentage values

Number of decimals. A non-negative integer of maximum 9 digits.

f

The logical expression that requests the filter population

Logical expression 

Write the logical expression that specifies the filter of the graph.

The filter will be a subset of the 'universe' attribute in the <xtgroup> element.

filtertext

A template text expression to generate the text associated with the logical expression

text 

Write the text that decribes the filter you have for the xt

If no filtertext is given there are two options: 1) When the filter (the 'f' attribute) is a BASIC logical expression (not combined from more questions) the standard template content will be that the generated text will be the text describing the right hand side of the logical expression (answer codes). 2) When the filter is a more COMPLEX logical expression, the logical expression will be shown as a filtertext. Known issue: The filtertext ONLY works on question level, NOT on a global level (xtgroup)

max

Maximum shown in the value axis

Specify an integer defining the maximum

min

Minimum shown in the value axis

Specify an integer defining the minimum

pptslidelayout

Specifies whether you want a standard PowerPoint layout (Title and content) or two charts on one slide (Two content)

ppLayoutchart 

Standard Title and content slide

ppLayoutTwoObjects 

Will show this in the two content slide

qno

Should the question box number (name) appear as a suffix to the question text?

yes 

The question ID will be used.

no 

The question box number will NOT appear.

sequential 

The question box sequential number in the questionnaire will appear.

The system will follow the template

rolldir

The direction of the rolling to be performed, if any.

asc 

Rolling is ascending

desc 

Rolling is descending

Example: Question 9.A is a brand awareness question, where the last code is 'Don't know'. The expression <g a='\9.A' b='ALL' rollexclude='1' rollend='5' rolldir='desc' /> will show the TOP (rolling='yes' rolldir='desc') 5 BRANDS (rollend='5') at the TOTAL LEVEL and the 'don't know any brand' (rollexclude='1') will also be shown at the bottom of the graph. Known issue: The filtertext ONLY works on question level, NOT on a global level (xtgroup)

rollexclude

Should a number bottom bars be excluded from the rolling prosess?

Specify the number of rows that shall be excluded from the rolling. Normally this will apply to 'don't know' and 'other' alternatives. These will come additionally to the numbers specified in the rollend attribute.

Example: Question 9.A is a brand awareness question, where the last code is 'Don't know'. The expression <g a='\9.A' b='ALL' rollexclude='1' rollend='5' rolldir='desc' /> will show the TOP (rolling='yes' rolldir='desc') 5 BRANDS (rollend='5') at the TOTAL LEVEL and the 'don't know any brand' (rollexclude='1') will also be shown at the bottom of the graph.

rollend

The top N bars of each graph block will be shown.

Specify the n numbers of rows that shall be shown in the graph ("Show me the TOP N brands").

Example: Question 9.A is a brand awareness question, where the last code is 'Don't know'. The expression <g a='\9.A' b='ALL' rollexclude='1' rollend='5' rolldir='desc' /> will show the TOP (rolling='yes' rolldir='desc') 5 BRANDS (rollend='5') at the TOTAL LEVEL and the 'don't know any brand' (rollexclude='1') will also be shown at the bottom of the graph.

rolling

Should the values be sorted in the graph

yes 

Rolling (sorting) should be performed

no 

Rolling (sorting) should NOT be performed (default value)

sat

Are <sat> definitions to be applied for this graph?

yes 

The <sat> definitions will be shown both in bar texts, legends and question texts.

no 

The <sat> definitions will NOT be shown.

break 

The <sat> definitions will ONLY be shown in LEGENDS.

row 

The <sat> definitions will ONLY be shown both in BAR TEXTS and QUESTION TEXTS.

seriesdoc

Documents the size of the series for categorical questions.

no 

No, will not show series documentation

yes 

Yes, will show series documentation. Default value is number of interviews

int 

Number of interviews

pop 

Weighted population

The series documentation will show together with the chart legend.

seriesin

Swap row/columns in the analysis: Specify which series (a='\xx' or b='\xx') that should be analyzed at the first level.

Analyze series A first

Analyze series B first

show

The sequence of show attributes in the bar blocks.

int 

Non weighted number

pctfilt 

Filter percentage

pcthori 

Horizontal percentage

pctvert 

Vertical percentage

pop 

Weighted population

Note: Average will be shown for all quantitative values

stacked

Should stacking be made?

yes 

Graphs will be stacked

no 

Graphs will NOT be stacked

step

Specify the steps for the gridlines in the chart.

An integer being the step value

If you specify e.g. a step value of 20, and a max value of 100, there will be a gridline for every 20 in the chart (20,40,60 and 80)

title

Overrides the chart title.

Text 

Specify the text you want to override the title.

Note: The title will also override any filters or basedocumentation that is in together with the original title

titlein

Will show the chart title and any documentation inside the chart itself. Else it will be shown as the PowerPoint SLIDE title

no 

If No or not specified, the title will be shown in the SLIDE title.

graph 

Specify if you want the title to be presented in the graph, not in the PowerPoint slide title.

This may be useful if you want to use the Slide title to comment the chart.

type

Specifies the chart type to be produced

hbar 

Horizontal bar chart

pie 

Pie chart

profile 

Profile bar chart

vbar 

Vertical bar chart

line 

Line

profilemid 

Profile bar chart

universe

The logical expression that defines the global filter for the graph

Logical expression 

Write the logical expression that specifies the universe of the graph.

If combined, the filter will be a subset of the 'universe' attribute in the <xtgroup> element.

values

Should values be displayed within the graph?

yes 

Values will be shown within the graph

no 

Values will NOT be shown within the graph

wset

A weighting set id to be applied in this graph.

name 

Insert the weight set ID (name) for the weight set to be applied


< xt > - Start a cross table or define <xt> attributes

To top
Its purpose is to start a cross table by <xt> or to define attributes to be inherited by subsequent <xt>'s. If only attributes are to be defined the tag name must be <xtgroup>

a

Defines the row margin of the cross table

address_expression 

Specify the number of adresses to be shown

Example: The adress expression a='\4[1,,4,-,+]' will show a table of question 4. The first four rows, no answer and sum of answers will ble shown. (Link: a.htm)

acc

Defines an address or address vector to be applied as a table accumulator

accumulator_adress 

Define a quantitative variable

Example: If question 4 is age groups and question 2 is age (quantitative) the following expression <xt a='\4' acc='\2.1'/> will show the age average in the different categories. (Link: acc.htm>

autocode

Ascending or Descending will make an address vector of all occurencies of a numeric data type (F) in the a attribute and the values will appear ascending or descending in the table. The H address component may be present in the a attribute in order to limit the value vector.

ascending 

The quantitative question will be shown in ASCENDING order

descending 

The quantitative question will be shown in DESCENDING order

no 

The quantitative question will NOT be autocoded.

Example: Question 2 is age. The table <xt a='\2' autocode='descending'/> will show the frequency of all age groups in the survey in a descending order. AUTOCODE IN INTERVALS? (Link: autocode.htm)

b

Defines the column margin of the cross table (break).

address_expression 

Specify the number of adresses to be shown

Example: If question 5 is household income, question 1 is gender and 4 is age groups, the table the following expression <xt a='\5' b='ALL \1 \4'/> will show the household income in the rows, and TOTAL; gender and age in table columns. (Link: b.htm)

bar

Specify yes or a list of column numbers if bars should be drawn in each table cell

n n...n 

Specify the column number(s) where bars should be drawn

Note: Not implemented in print, hence it works only in html. Furthermore, only one value is implemented. Example: If question 4 is age groups and question 5 is household income, the table the following expression <xt a='\5' b='ALL \4' bar='3'/> will show the results of column 3 also in a graphic representation. (Link: bar.htm).

basec

Default base for categorical values/computations. If given in the address this will take precedence.

all 

All respondents

Question asked

Question answered

� 

Question exist

baseq

Default base for quantitative values/computations. If given in the address this will take precedence.

all 

All respondents

Question asked

Question answered

� 

Question exist

baserows

When bases that effect the show attributes change within the table there is an option to determine whether these bases should be printed or not.

int 

Number of cases base row

market 

Market size base row

none 

No base rows wanted

pop 

Population base row

sum 

Accumulated sum base row

Leave it to the system to decide

Note: The value none is exclusive to the others, all other combinations are possible.

cdtext

User defined text that should appear in the CD cell (Cell Documentation cell) for this table. If not present this cell will contain standard generated text.

text 

Specify text to be shown in the CD cell

chi2

Should chi2 test be performed? And at which level?

yes 

By specifying chi2='yes' the test level is determined by the table filter number of cases as follows: 10% when cases = 1-300, 5% when cases = 301-1000, 2,5% when cases = 1001-4000, 0,5% when cases = 4001-20000 0,1% when cases = 20001 or more

Specify the Chi2 test level for this table. The value is restricted to values ranging between 0 and 50. The value 0 means that no Chi2 test is performed.

Can be combined with two other attributes: <wchi2> and <chi2calc>.

chi2calc

Should chi2 test be performed with normal (symmetric) or single class calculation?

normal 

Chi2 values will be calculated with normal (symmetric) calculation

singleclass 

Chi2 values will be calculated with single class calculation

Must be combined with the <chi2> attribute, and can be combined with the <wchi2> attribute.

deciacc

The number of decimals to be used for accumulated sum values

Number of decimals. A non-negative integer of maximum 9 digits.

deciavg

The number of decimals to be used for average and fractile values

Number of decimals. A non-negative integer of maximum 9 digits.

decichi2

The number of decimals to be used for Chi2 statistics

Number of decimals. A non-negative integer of maximum 9 digits.

decidev

The number of decimals to be used for quantitative deviation measures

Number of decimals. A non-negative integer of maximum 9 digits.

decipct

The number of decimals to be used for percentage values

Number of decimals. A non-negative integer of maximum 9 digits.

decipop

The number of decimals to be used for population values

Number of decimals. A non-negative integer of maximum 9 digits.

decimalsymbol

 

text 

excludepair

Specify a series of two-letter combinations representing the pairs of columns that normally would be tested by the T test, but will now be excluded from the test.

ab ac ad bd ... 

Specify combinations that are NOT going to be tested against each other in the T-test

f

The logical expression that requests the filter population

Logical expression 

Write the logical expression that specifies the filter of the xt.

The filter will be a subset of the 'universe' attribute in the <xtgroup> element.

filterdoc

Should the filter box be printed?

yes 

Filter documentation will be printed

no 

Filter documentation will NOT be printed

Yes if the f attribute is present, No if f attribute is not present

filtershow

Which figures should be present within the filter box:

int 

Number of cases

pop 

Population size

share 

Population size as percentage

none 

No figures to be displayed

filtertext

A template text expression to generate the text associated with the logical expression

text 

Write the text that decribes the filter you have for the xt

If no filtertext is given there are two options: 1) When the filter (the 'f' attribute) is a BASIC logical expression (not combined from more questions) the standard template content will be that the generated text will be the text describing the right hand side of the logical expression (answer codes). 2) When the filter is a more COMPLEX logical expression, the logical expression will be shown as a filtertext.

includepair

Specify a series of two-letter combinations representing the pairs of columns that NORMALLY would not be tested by the T test, but will now be included from the test.

ab ac ad bd ... 

Specify combinations that are going to be tested against each other in the T-test

introw

Should the number of cases row at table level be printed?

yes 

Number of cases will be shown

no 

Number of cases will NOT be shown

System will follow the template

lowcol

The minimum number of cases that a column should have in order to show any values in that column.

An integer being the minimum numberf of cases

pctsign

Which percentage figures should have the % suffix? Specify the figures you want to show with the % suffix:

filt 

filter percentage

pcthori 

horizontal percentage

pctvert 

vertical percentage

hsum 

horizontal sum

vsum 

vertical sum

Example: <xt a='\4' showc='pctvert int pop pcthori' pctsign='pctvert pcthori'/> will show both horizontal and vertical percentages with a % suffix

poprow

Should the table top population row be displayed?

yes 

There WILL be a table top population.

no 

There will NOT be a table top population

The system will follow the template

qno

Should the question box number (name) appear as a suffix to the question text?

yes 

The question ID WILL appear.

sequential 

The question box sequential number in the questionnaire will appear.

no 

The question box number will NOT appear.

The system will follow the template

rolldir

The direction of the rolling to be performed

asc 

Rolling is ascending

desc 

Rolling is descending

Example: Question 9.A is a brand awareness question, where the last code is 'Don't know'. The expression <xt a='\9.A' b='ALL \1.A \4.A' rollitem='1' rollvalue='pctvert' rolldir='desc' rollexclude='1' rollend='5' /> will show the TOP (rolling='yes' rolldir='desc' rollvalue='pctvert') 5 BRANDS (rollend='5') at the TOTAL LEVEL (rollitem='1') and the 'don't know any brand' (rollexclude='1') will also be shown at the bottom of the table.

rollexclude

Should a number bottom bars be excluded from the rolling prosess?

Specify the number of rows that shall be excluded from the rolling. Normally this will apply to 'don't know' and 'other' alternatives. These will come additionally to the numbers specified in the rollend attribute.

Example: Question 9.A is a brand awareness question, where the last code is 'Don't know'. The expression <xt a='\9.A' b='ALL \1.A \4.A' rollitem='1' rollvalue='pctvert' rolldir='desc' rollexclude='1' rollend='5' /> will show the TOP (rolling='yes' rolldir='desc' rollvalue='pctvert') 5 BRANDS (rollend='5') at the TOTAL LEVEL (rollitem='1') and the 'don't know any brand' (rollexclude='1') will also be shown at the bottom of the table.

rollend

The top N rows of each table block will be shown.

Specify the n numbers of rows that shall be shown in the table ("Show me the TOP N brands").

Example: Question 9.A is a brand awareness question, where the last code is 'Don't know'. The expression <xt a='\9.A' b='ALL \1.A \4.A' rollitem='1' rollvalue='pctvert' rolldir='desc' rollexclude='1' rollend='5' /> will show the TOP (rolling='yes' rolldir='desc' rollvalue='pctvert') 5 BRANDS (rollend='5') at the TOTAL LEVEL (rollitem='1') and the 'don't know any brand' (rollexclude='1') will also be shown at the bottom of the table.

rollitem

Which column or pair of columns determine the rolling prosess?

Specify column number.

Example: Question 9.A is a brand awareness question, where the last code is 'Don't know'. The expression <xt a='\9.A' b='ALL \1.A \4.A' rollitem='1' rollvalue='pctvert' rolldir='desc' rollexclude='1' rollend='5' /> will show the TOP (rolling='yes' rolldir='desc' rollvalue='pctvert') 5 BRANDS (rollend='5') at the TOTAL LEVEL (rollitem='1') and the 'don't know any brand' (rollexclude='1') will also be shown at the bottom of the table.

rollvalue

Which cell show attribute should be used for the rolling?

Avg 

Average

Index 

Categorical index

Int 

Non weighted number

Pcthori 

Horizontal percentages

Pctvert 

Vertical percentages

Pop 

Population sizes

Sum 

Accumulted sum

Population for categorical tables, average for quantitative tables

Example: Question 9.A is a brand awareness question, where the last code is 'Don't know'. The expression <xt a='\9.A' b='ALL \1.A \4.A' rollitem='1' rollvalue='pctvert' rolldir='desc' rollexclude='1' rollend='5' /> will show the TOP (rolling='yes' rolldir='desc' rollvalue='pctvert') 5 BRANDS (rollend='5') at the TOTAL LEVEL (rollitem='1') and the 'don't know any brand' (rollexclude='1') will also be shown at the bottom of the table.

sat

Are <sat> definitions to be applied for this table?

yes 

The <sat> definitions will be shown both in rows, columns and question texts.

no 

The <sat> definitions will NOT be shown.

break 

The <sat> definitions will ONLY be shown both in COLUMNS.

row 

The <sat> definitions will ONLY be shown both in ROWS and QUESTION TEXTS.

showbreak

Should the cases and population table top row show horizontal percentages?

Show horizontal percentages in the top row.

no 

Do NOT show horizontal percentages in the top row.

showc

The sequence of show attributes in categorical row blocks.

chi2d 

Chi2 distribution value (level) connected to the Chi2 statistic

chi2s 

Chi2 statistic (weighted and normed at case level)

gamma 

Gamma statistic

index 

Categorical index

int 

Non weighted number

pctfilt 

Filter percentage

pcthori 

Horizontal percentage

pctvert 

Vertical percentage

pop 

Weighted population

showcall

The sequence of show attributes in ALL columns of categorical row blocks.

chi2d 

Chi2 distribution value (level) connected to the Chi2 statistic

chi2s 

Chi2 statistic (weighted and normed at case level)

gamma 

Gamma statistic

index 

Categorical index

int 

Non weighted number

pctfilt 

Filter percentage

pcthori 

Horizontal percentage

pctvert 

Vertical percentage

pop 

Weighted population

showq

The sequence of show attributes in quantitative row blocks.

avg 

Average

fNNN 

The NNN% fractile. The attribute values fNNN may take the values from f0 through f100 with the following interpretations: where f0=minimum value, f25=lower quartile, f75=upper quartile, f100=maximum value and f50=median value.

index 

Quantitative index

int 

Non weighted number

mode 

The 'most frequent number'

pctfilt 

Filter percentage

pcthori 

Horizontal percentage

pcthsum 

Horizontal sum percentage (Share of quantity)

pctvert 

Vertical percentage

pop 

Weighted population

std 

Standard deviation

ste 

Standard error of mean

sum 

Quantity sum

showqcall

The sequence of show attributes in ALL columns of quantitative row blocks.

avg 

Average

int 

Non weighted number

pctfilt 

Filter percentage

pcthori 

Horizontal percentage

pcthsum 

Horizontal sum percentage (Share of quantity)

pctvert 

Vertical percentage

pop 

Weighted population

std 

Standard deviation

ste 

Standard error of mean

sum 

Quantity sum

summary

This option requests that only the calculated scaling rows are printed when the scaling feature is defined as part of the address.

yes 

ONLY the calculated scaling values will show

no 

The quantitative question will show as normal

Example: Question A1 is a grd question with a scale from 1 (disagree) to 7 (agree). The follwing table <xt a='\A1[1,,7<>]' summary='yes' showq='avg'/> will show ONLY the average of the scaling, where the codes has been given the code value

textcheck

When a scalar Q component of an address is composed of several question boxes or when a scalar S component through several question boxes are detected there is an option on checking whether the text elements should all be equal to allow the analysis to be done.

yes 

All texts have to be equal

no 

Texts can be different

Example: Three questions \Q1 (First mentioned), \Q2 (Second mentioned) and \Q3 (Mentioned later) are brand awareness questions covering the same brands. To see total awareness, the table <xt a='\Q1:Q3' textcheck='yes'/> will produce the total awareness of the different brands, and at the same time check that all rows ar the same.

title

Using this attribute will have these affects: The text will appear at the top of the table heading box The text will appear as an index label to the Live Table application. If NEWTABLE=NO then this text will appear at row block text to this table section

text 

Specify your table title

ttest

Should ttest be performed? And at which level?

yes 

By specifying ttest='yes' the test level is determined by the table filter number of cases as follows: 10% when cases = 1-200, 5% when cases = 201-400, 2% when cases = 401-800, 1% when cases = 801-1600, 0,5% when cases = 1601-3200, 0,2% when cases = 3201-6400, 0,1% when cases = 6401 or more

Specify the ttest test level for this table. The value is restricted to values ranging between 0 and 50. The value 0 means that no ttest test is performed.

universe

The logical expression that defines the global filter for the cross table

Logical expression 

Write the logical expression that specifies the universe of the cross table.

If combined, the filter will be a subset of the 'universe' attribute in the <xtgroup> element.

wchi2

Should the weighted Chi square test be performed?

yes 

The weighted Chi2 test will be performed

no 

The weighted Chi2 test will NOT be performed

Must be combined with the <chi2> attribute, and can be combined with the <chi2calc> attribute.

wset

A weighting set id to be applied in this table.

name 

Insert the weight set ID (name) for the weight set to be applied

zerocell

A character string to be shown in cells with no cases.

Character 

Insert a character to be shown in cells with no cases. (Default = dash ('-')

zerorow

Rows that have no cases may be left out

yes 

Rows with no cases will be shown

no 

Rows with no cases will be left out and NOT shown


< dig > - <dig> Create a Dig analysis

To top
Its purpose is to create a target group analysis by using the <dig> or to define attributes to be inherited by subsequent <dig>'s. If only attributes are to be defined the tag name must be <xtgroup>

show

Which values are to be shown in the Dig analysis

desc 

Question and category text

addr 

Question ID

vert 

Coverage in target group

nvert 

Coverage in non-target group

size 

Total coverage

chi2d 

Significance level

hori 

Share of category in target group

nhori 

Share of category in non-target group

chi2 

Value of Chi2 statistic

gamma 

Value of gamma

index 

Index value

Specify the values that will be shown in a space separated list. Default values are: 'desc addr vert nvert size chi2d'

a

Specify the question(s) you want to analyze. Set a='\' to get the complete questionnaire as-is, or use an expression to get parts of it.

address_expression 

Specify a number of questions to analyze

You may also use the attribute <addr>. It does the same. Best practice is often to specify topbox or bottombox in grid questions, in order to make the groups more robust, e.g. a='\Q5[0:6,9:10]' will only show the topbox and bottombox of a NPS question from 0 to 10.

f

Filter. Specify the target group you want to analyze.

Logical expression 

Use a normal logical expression

If you combine the filter definition with the use of the <Universe> attribute, the filter (or a part of it) must be a real subset of the universe definition

include

Should positive or negative significant findings be shown

Show only positive significant findings

Show only negative significant findings

pn 

Show both positive and negative significant findings

Standard is showing both positive and negative findings

limitchi2

At what Chi2-significance level should the report stop showing results

Specify a Chi2-level

A standard limit of 4 is implemented, i.e. all significant findings that are listed in the report are significant at a 95% level.

limitgamma

At what gamma-significance level should the report stop showing results

Specify a gamma-level

The standard report uses a limit of 0 as the limitation, showing only positive values.

limitindex

At what index level should the report stop showing results

The standard report uses a limit of 0 as the limitation, showing only values above 100.

limitnumber

How many findings should be presented in the Dig-report.

Specify the number of categories to be shown

Default is to show all categories that satisfies the significance criterias

decimalsymbol

Any symbol/text to use as decimal symbol

text 

Specify a decimal symbol: comma(,) or dot(.).

decipct

The number of decimals to be used for percentage values

Number of decimals. A non-negative integer of maximum 9 digits.

decichi2

The number of decimals to be used for Chi2 statistics

Number of decimals. A non-negative integer of maximum 9 digits.

lang

In which language will the standards texts be shown?

NN 

Specify the language by the language two letters.

Example: lang='no' will show the table in norwegian. Unless other is specified, default language will be English

filtertext

A template text expression to generate the text associated with the logical expression defined in the target group (<a>-attribute)

text 

Write the text that decribes the filter you have for the xt

If no filtertext is given there are two options: 1) When the filter (the 'f' attribute) is a BASIC logical expression (not combined from more questions) the standard template content will be that the generated text will be the text describing the right hand side of the logical expression (answer codes). 2) When the filter is a more COMPLEX logical expression, the logical expression will be shown as a logical expression.

filterdoc

Should the filter box be printed?

yes 

Filter documentation will be printed

no 

Filter documentation will NOT be printed

filtershow

Which figures should be present within the filter box:

int 

Number of cases

pop 

Population size

pctfilt 

Population size as percentage

Specify using a space separated list. Default values are :'int pop pctfilt'

rolling

Which value should be used for the sorting?

chi2 

Sort by Chi2 statistics value

hori 

Sort by hare of category in target group

vert 

Sort by coverage in target group

gamma 

Sort by gamma-value

index 

Sort by index-value

The sorting value does not neccesarily need to be a part of the table. Default value is sorting by Chi2 value.

rollsign

Should the list be sorted by using absolute values or actual values.

yes 

Sort by absolute values

no 

Sort by actual values

Sorting by absolute values will show the positive and negative findings mixed in the sorting order

universe

The logical expression that defines the global filter for the dig analysis

Logical expression 

Write the logical expression that specifies the universe of the dig analysis.

If you combine the filter definition with the use of the <Universe> attribute, the filter (or a part of it) must be a real subset of the universe definition