AskHYS.net Q x Q – Online Data Query System Instructions

What is Q x Q?

QXQ is a query system on AskHYS.net, a website for the Washington State Healthy Youth Survey. Using the QXQ you can find answers to questions from the Healthy Youth Survey, question by question---Q by Q, Q X Q. The “X” or “by” also means that you can run crosstabs---that is, create a table with, for instance, smoking in the rows “by” drinking in the columns. You’ll see what we mean….

What Can You Do with the Q x Q?

The QXQ allows you to run frequencies and crosstabs with Healthy Youth Survey data.

• Frequencies: the number, or frequency, of specific answers to specific questions. For instance, the number of students who reported smoking in the past 30 days.
• Crosstabs: the relationship between two questions, or variables---crossing one question by the second question. For instance, smoking X drinking will tell you if kids who smoke also drink.

What Survey Data Can You Use?

Anyone can use the state sample, and most county data, for every year (2002, 2004, 2006, and 2008) and every grade in which the survey was implemented (6th, 8th, 10th, and 12th grades).

TWO IMPORTANT NOTES:

1. You can run an analysis on only one year and one grade at a time.
2. State, county, and ESD results are available to everyone. But access to school district and school building data requires that you log on with a password:

Logging On

• Do you need a password? Click the Log On at AskHYS.net home page. Then select “Need help getting access to district or building results?” for an explanation of password access.
• If you have a password and are able to log on, the Q x Q will display a drop down menu with the locations (districts or buildings) that you have access to.
• Some districts (for instance, very small ones) have only school building data. If you try to run a district report and get an error message, or get no data, try to run a building report.

Does my school/school district have data?

• Not all locations participate in the HYS every year. For some schools participation rates are too low to produce reliable results.
• To find out what results are available, go to the Available Results link at the top of the Home page.

What Survey Questions are Available

The questions available on the Q x Q depend on the year of the survey, and the grade level.

• Some survey questions are not asked every year, or change slightly from one year to the next.
• Many survey questions are only on the secondary grade survey (8th, 10th and 12th grades). There are two versions, Form A and Form B. In a classroom, half of the students take Form A, and half take Form B.
• The 6th grade survey, Form C, is much shorter.

Getting Started with Your Analysis

STEP 1: Choose Initial Analysis Variables

Using the drop down boxes at the top of the Q x Q, select year, grade, gender and geography (or location):

STEP 2: Select Survey Questions

First: Click the drop-down menu “Select a survey data category…”

Next: Select a category, and then a sub-category

• Example: Say you are interested in how easy kids think it is to get access to alcohol. Scroll down the list to the Alcohol topics, and then select the sub-topic, “Access”.

The survey questions that relate to alcohol access will appear in the drop down menu below.

NOTE: The numbers in parenthesis on the left correspond to the dataset variable name in parentheses. Those numbers can help you navigate in the Survey Question Crosswalk. (See FAQ about finding survey questions

When you click on a question title, in this case, Access to Alcohol, the “Item response category preview” box to the right will have the survey question and its response options. This is a good way to verify that you are looking at the question you really want.

TIP: You can also “right click” on a question title to bring up a pop-up box. Click on “See all years’ descriptions” and it will provide the question, response options and the years it was asked.

STEP 3: Drag and Drop Questions into Analysis Boxes

Dragging and Dropping
Click and hold on the question you want to analyze, and drag it into the analysis box labeled “Drop first variable (row) here….”

OR:

You can also right click on the question and you’ll get a pop up with the option to “Include in analysis”. Then select “row” or “column”.

STEP 4: Choose Response Options

The box to the right will show you the response options for the question. “Surveyed” will give you all of the options that were available to the respondent. “Collapsed” reduces the response options into only two responses

TIP: Why would you use the “collapsed” categories? By combining the answers of two or more responses, you also combine the number of respondents in the “collapse” categories. This will be important when you run crosstabs because you MUST have a minimum number of respondents in each cell of your results table. [There is more about “cells” in the next section on crosstabs.]

For [M10] Access to Alcohol:

Collapsed	Surveyed

STEP 5: Run Query

Click the “run query” button at the bottom right side of the screen.

EXAMPLE: In this example, we selected: year 2008, grade 8, location State WA, and the question [M10] Access to Alcohol. For the response options we chose “surveyed”.

The results include the percentage of students who chose this option, the 95% confidence interval and the number of respondents for each response option. For more information about confidence intervals and how to use them, see the Frequently Asked Questions section of this document.

Changing to a New Analysis: Drag, Click, or Reset

To change to a different question:

• First remove the original question from the analysis box, either drag and drop back up to the List of Questions box, OR
• Right click the question, and then click on “Remove from analysis”, OR
• Hit the Reset button at the bottom of the Q x Q on the left side.
• This will remove your survey questions from the analysis boxes, but will not change year, grade, or location.

To use the same question but for a different grade:

• Return to the top of the Q x Q menu, and change the grade level. The other settings will stay the same.

EXCEPTION: If you want to run 6th grade, you will have to re-select your question.

Running Crosstabs with HYS Data

What is a Crosstab?

A crosstab allows you to see the relationship between two different questions from HYS. In the example below, we see the relationship between drinking and smoking: 95.4% of youth who reported NO drinking on any day(s) in the past month did not smoke (cell 1), but 35.5% who drank also report smoking (cell 4).

Requirements for Crosstabs

**NOTE**: The squares that have results in them are the “cells”, and the number of respondents in each cell is called the “n”, or also the “cell size”.

Cell Size

• For state level analysis you must have 5 or more respondents in each cell.
• For sub-state level analysis you must have 10 or more respondents in each cell.

If you don’t have enough respondents per cell, you will receive the following error message:

“At least one cell in the results table contained a count of less than 10.“ Output is suppressed.”

Cell size limitations can be frustrating. In the example above, we were using the state sample, so the n’s are large. If you were running a crosstab for a small district or school the number of responses in each cell would be much smaller, so you may not get a report. If your results are suppressed because of cell size, try this:

• Select the “collapsed” answer option rather than the full set of “surveyed” options.
• Re-run your crosstab at school district rather than school, or county rather than school district.

Survey Form: A, B, or C

The secondary school version of the survey has two distinct versions: Form A and Form B. In each classroom, half the students get Form A, and half get Form B. Most of the survey questions are either on Form A or Form B. For a crosstab, both questions have to come from the same form---you can’t cross a question that is only on Form A with another question that is only on Form B.

For example, you cannot cross current cigar smoking (only on form B) with current methamphetamine use (only on form A). You will receive the following error message:

“No surveys contained responses to all the selected variables”.

About 35 questions are identical on both versions. For those questions, like 30-day cigarette use or 30-day alcohol use, you can run crosstabs with any question on either survey. This is important because Form A has most of the risk and protective factors, and Form B has most of the physical and mental health questions.

Running a Crosstab: STEPS 1-3

To run a crosstab, you use the menus the same way that you would for a frequency, but now you chose two questions instead of one.

1. STEP 1: Drag or insert your first question in the “Drop first variable (row) here” analysis box.
2. STEP 2: Drag or insert your second question in the “Drop second variable (column) here” analysis box.
3. STEP 3: Select response options for both questions.

In this example, we have current alcohol drinking as the first variable, and current cigarette smoking as the second variable. For both questions we have selected the collapsed response options.

Running a Crosstab: STEP 4

Hit the Run Query button.

Interpreting Crosstab Results

Read your results by the row, not by the column. Notice that the “Total” for each row is 100%. So this means that the top row has results for all of the students (100%) who reported no alcohol drinking.

• Top row, Cell 1: 95.4% who did not drink alcohol, did not smoke cigarettes
• Top row, Cell 2: 4.6% who did not drink alcohol, did smoke cigarettes

The bottom row has results for all of the students who reported at least one day of drinking.

• Bottom row, Cell 3: 64.5% who drank alcohol, did not smoke cigarettes
• Bottom row, Cell 4: 35.5% who drank alcohol, did smoke cigarettes

You can’t read the columns in the same way!

If you want to know how many smokers drink, re-run your crosstab. Select:

• Current Cigarette Smoking as your first question, so that it will be in the Rows, and
• Current Alcohol Drinking as your second question, so it will be in the columns.

TIP: Another way to say this, the report we ran is about student non-drinkers and student drinkers. And we are asking if they smoke. If you want to find out if the smokers and non-smokers drink, well you have to re-run the crosstab, and put smoking in the rows, and drinking in the columns.

TIP: The Table Preview in the upper right hand corner can give you an idea about how your analysis will turn out. (The table uses variable number rather than the whole item name.)

In this example, [D20] Current Alcohol Drinking will appear in the rows, and [D14] Current Cigarette Smoking will be in the columns. So your results will tell you how many drinkers smoke cigarettes.

Testing for Significant Differences

When an analysis you have done is interesting or important for your work, you may want to discover if the differences you have found between two groups is “statistically significant”. This is important when you are using data from a sample, and want to generalize to a larger population.

If the difference between two groups (say, the difference in smoking rates between non-drinkers and drinkers in the example above) is “statistically significant” it means that we feel confident that the difference is not due to chance alone. That confidence is expressed as a probability. A commonly used probability, 95%, can be interpreted as “we are 95% confident that the difference is significant”, or that chance could explain the difference only 5% of the time.

You can assess significance using confidence intervals, but it is more precise to determine significance mathematically, and you can do it easily. Go to http://www.hys.wa.gov/Reporting/Default.aspx, scroll down to 3rd bullet under Information and Tools, “Excel Tool for Determining Statistical Significance”. All you have to do is enter your data---the percentage, and the plus-or-minus (the margin of error, ±) into the boxes.

EXAMPLE:

To test for differences in drinking soda between 10th and 12th graders in 2008, statewide:

Year: 2008	Grade: 10th	Gender: Both	Location: State
First Variable: Nutrition – Junk Food - [H09] Soda Drinking (collapsed)
Results = 2 or more sodas for 10th grade: 15.3% (±1.7)

Year: 2008	Grade: 12th	Gender: Both	Location: State
First Variable: Nutrition – Junk Food - [H09] Soda Drinking (collapsed)
Results = 2 or more sodas for 12th grade: 14.7% (±1.9)

Interpretation: P-value is greater than 0.05, so there is no difference in drinking 2 or more sodas between 10th and 12th graders, in 2008 statewide.

Test for differences in drinking soda between 10th grade males and females in 2008, statewide:

Year: 2008	Grade: 10th	Gender: Male	Location: State
First Variable: Nutrition – Junk Food - [H09] Soda Drinking (collapsed)
Results = 2 or more sodas for 10th grade: 20.5% (±2.3)

Year: 2008	Grade: 10th	Gender: Female	Location: State
First Variable: Nutrition – Junk Food - [H09] Soda Drinking (collapsed)
Results = 2 or more sodas for 12th grade: 10.5% (±1.9)

Interpretation: P-value is less than 0.05, so 10th grade males are more likely to drink 2 or more sodas compared to 10th grade females, in 2008 statewide.

If you want to learn more about statistical significance, including how to interpret the confidence intervals in your QXQ results, go to the Frequently Asked Questions in this handout, or to page 36 in the HYS training workbook, at http://www.hys.wa.gov/Workshops/2008/HYS09WorkbookFinal.pdf.

Resources

https://fortress.wa.gov/doh/hys

• Background information and technical documentation
• Reports of Results for the state, counties and sub-populations like gender and race/ethnicity
• Copies of the survey instruments
• A manual on how to conduct HYS analysis
• An analytic report with trend results

http://www.hys.wa.gov

• Information on HYS registration
• History of school participation
• Survey administration and training materials
• Tool for testing significance, go to the website
• Reports, including a history of Risk and Protective Factors used on the survey
• Training manual for using Healthy Youth Survey reports

http://www.AskHYS.net

• A list of Joint Survey Planning Committee contacts
• Pre-formatted reports with questions arranged by topic

For help with the AskHYS.net Q x Q, contact: Susan Richardson: sue.richardson@doh.wa.gov

Frequently Asked Questions

Did my School Participate?

The results are displayed in the following order: state, ESD, county, district, school building and then consortia*. School district and building results are arranged by county, so the districts in Adams County are first and those in Yakima County are last.

Select the year of results you want and scroll or search to find your ESD, county, district or school:

• If the grade and location say “Yes”, then results are available.
• If the grade and location say something else, use the key on the Available Results page to determine why they are not available.
• If you are looking for district and county results and it says “Too few schools” or “Only 1 school”, scroll down to the school building results to see if they are available.

Why do I Have to Start Over Every Time I Want to Change my Analysis to the Sixth Grade?

Each time you chose a different survey year or select 6th grade rather than a secondary school grade, the available questions will change.

How do I Find the Survey Question I Want?

There are a lot of questions on HYS and many categories and sub-topics to choose from. If you are having difficulty finding the question you want, try using the Survey Question Crosswalk. You can find that this website: https://fortress.wa.gov/doh/hys/SurveyQuestions.htm. We will soon add that to the AskHYS website.

The Master tab of the crosswalk includes the actual HYS question, the sub-topic, and the title. You can search for a specific word to find your question and where to find it in the drop downs.

The columns on the right of the spreadsheet include the variable names for 2008, 2006, 2004, and 2002. These columns not only tell you the variable names, but let you know if the question was asked that year. If a question title is missing for a specific year that means the question was not asked then.

There are also year tabs: 2008, 2006, 2004, and 2002:

• Each of these tabs provides you with the actual HYS question and response options.
• It also gives you the variable name, where it is located in the Report of Results from RMC, and which form of the survey the question was on.

TIP: If you have the question you want, but simply want to know if it is available in other years, in the QXQ dropdown list you can “right click” on a question title to bring up a pop-up box. Click on “See all years’ descriptions” and it will provide the question, response options and the years it was asked.

Knowing which form the question was on can be helpful in determining:

• Which HYS questions are core (Item numbers are included in the cells for both Form A and Form B)
• Which questions are available for 6th graders (Item number is included in the cell for Form C)
• Which results are only available for 6th grade (there is an x in both the Form A and Form B cells, and an item number in the Form C cell).

What are “Confidence Intervals” and How should I Use Them?

AskHYS results include a ± number after each item estimate—this number is a confidence interval. A confidence interval accounts for the fact that the reported value is probably a little different than the true value for all of the students. A 95% confidence interval, for example, means that we are 95% confident that the true value is within the ± range. Confidence intervals are important when you generalize results to a larger population.

What is “Statistical Significance”, and When Does It Matter?

Statistical significance means that the probability that differences in results are not due to chance alone. When using 95% confidence intervals, a difference between two groups is considered statistically significant if chance could explain it only 5% of the time or less.

Confidence intervals can help you quickly determine significant differences, but there are more precise ways to determine significance. We will be showing you a tool later that can help with this. Also, assistance determining statistical significance is available from many sources including the local health department, the local ESD, JSPC agencies, or the Internet.

Can I Compare my QXQ Results to Other Reports?

You can compare the Q x Q results with the 2008 Statewide Grade 8 Report of Results (from RMC), available online at: https://fortress.wa.gov/doh/hys under Reports and Response Rates:

152.	If you wanted to get some beer, wine, or hard liquor (for example, vodka, whiskey, or gin), how easy would it be for you to get some?	State (n = 4,310)
	a. Very hard	36.2%	(± 1.9%)
	b. Sort of hard	26.9	(± 1.4)
	c. Sort of easy	20.8	(± 1.4)
	d. Very easy	16.1	(± 1.4)

Notice that there are some slight differences between the Q x Q results and the Report of Results. For Very Hard, the Q x Q results is 36.1% ±2.0 and in the Report of Results it is 36.2% ±1.9. These discrepancies can result from the way in which different statistical programs round numbers. If the discrepancy is small, it is not a problem.

To make sure the differences are just due to rounding, check the number of respondents for each response option. Notice that the Report of Results n is 4,310. If you add up the respondents from the Q x Q analysis (1,558 + 1,160 + 897 + 695) you also get an n of 4,310.

Why Can’t I Get Results because of the “Cell size”?

The cell-size rule is designed to protect the anonymity of students taking the HYS, and is required the Washington State Institutional Review Board (WSIRB).

More Examples

Frequencies by Year:

Frequencies by Grade:

Crosstabs by Gender:

Crosstabs by Race/Ethnicity:

Additional Crosstabs

Example bullying and depression:

Example eating breakfast and academic achievement (grades)