Frequently asked questions

Below is a list of frequently asked questions about the American Life Panel and its Data Pages. Should the answer to your question not be listed here, please just contact us at alp@rand.org.

How do I navigate to a particular survey or question?

Start off by clicking on Data in the navigation bar at the top of the screen. This will bring up a list of all available surveys. Next, to find more information for a survey, click on the link on the name of the survey. This will take you to the Survey Page. At the top of this page you will find a brief description of the survey, who the survey author(s) are and a link providing access to the response rate overview. Below that you will see a table of modules.

To go into an individual module, simply click on the name of the module. This will open the Module Page, which displays a table listing all the questions in the module. To further drill down to a particular question, just click on a question name. The resulting Question Page will provide you with the question specifics, such as the question label, question text, answer categories (if applicable), and any relevant notes. If available, an aggregate overview of the data associated with the question will also be shown.

I am lost. How do I get back to the overview of available surveys?

At the top of each Survey Page, Module Page , or Question Page you will find a set of navigation links that assist you with moving back to where you came from. For example, to get back to the main page after opening a particular survey, click on "All surveys". If you are in a particular module, clicking on the survey name will bring you back to the Survey Page. If you are viewing a specific question, clicking the module name will take you back to the Module Page.

How do I access the data?

Once you have registered as an ALP Data Pages user (click here to register), you can download data for an individual survey from its Survey Page. You will find the download link towards the bottom of the page below the "Data" header. You can also view data by selecting one or more questions in the Module Page of a survey, and then select "Show data" just below the questions table where it says "With Selected". This will open a window in which you can browse through the data and/or download it as a STATA or CSV file. Lastly, data aggregates for specific questions can be viewed on each individual Question Page.

Please note that if the status of survey data says "Under Embargo", this means that the data is not currently available to the public. Only survey authors who have registered as an ALP Data Pages user can access the data at this time. Surveys under embargo are highlighted in pink and include a lock symbol. If you click on the survey, you will see under "Description" the date the survey embargo will be lifted.

Can I construct my own dataset?

In addition to downloading datasets for particular surveys or for a sub-set of questions within a specific survey module, you can also construct your own dataset. For this purpose a so-called "question cart" is available to you. You can access your cart at any time by clicking on the question cart image located to the left of your username in the top right of the screen.

You can add questions to the cart in a variety of ways. One is to navigate to a specific Question Page and click the "add to cart" link (located at the far right of the question screen at the same height as the question name). Another option is to select one or more questions on a Module Page and select "Add to cart" next to "With selected". If you just completed a search and found one or more questions you would like to keep, you can opt to "check all" or individually put a check in one or more of the boxes in the far left column of the table. Next, choose "Add to cart" to the right of "With selected".

Once you have added questions to your cart, you can click on the question cart image. On the resulting page you can see what is currently in your cart and perform several functions on either specific or all questions in the cart:

What information is in the Demographics module I see for every survey?

All surveys come with a set of demographics variables, which can be found in the Demographics module of each survey. These variables contain the demographics information about each respondent as it was at the time he/she started the survey. More information about the demographic variables that are available and how they are collected can be found here.

Why does it sometimes take a while before I am prompted to download a data file?

For all surveys, a dataset is generated in real-time when you click on a STATA/CSV download link. This assures that you always receive the most recent data. For some of the larger surveys, the generating process takes a little more time. Please be patient. An alternative that may be faster (if you only need some of the variables within a survey) is to go to the Module page of a survey, check the specific variables you are interested in, and then select "Show data" just below the questions table where it says "With Selected". In the resulting window you will be offered an option to either download the data for the selected questions as a STATA or CSV file.

Why does my dataset download contain a CSV file saying something about truncation?

Sometimes the names of variables in a dataset exceed the 32 character limit imposed by STATA. For such variable names the generated STATA file will contain shortened variants. For each variable whose name was shortened, the original variable name can be viewed by viewing the characteristics of that variable using the following STATA command: 'char list shortenedname[]' (or 'char list' to view a list of all shortened variable names and their original counterparts). Additionally, a CSV file with a list of shortened names is also included in the downloaded zip file. Note that CSV files do not suffer from this limitation, and as such will always contain the full variable names.

STATA says that it can not open my dataset. What is wrong?

Sometimes the size of a downloaded dataset will be larger than the default allowance for Stata. In this case, please consult the information pertaining to your version of Stata. Generally, you will want to try one of the "set memory" commands in order to increase the size allowance and view the dataset.

STATA opens my dataset, but when I for example calculate the mean of a variable the result does not make sense at all. What happened?

On rare occasions the created STATA file may get corrupted during the generating or download process. If you suspect this is the case, just inspect the data in the data editor for a moment. Should the displayed values be garbled or not make sense in correlation to the variables to which they belong (e.g. a value of 32768 for a yes/no variable that should hold either 1 or 2), this may indicate that a problem has occurred. Please send us a message at alp@rand.org with details about the survey dataset (or set of variables from different surveys) that you were trying to download. We can then address the issue and provide you with the data you were attempting to obtain.

There are variables in my dataset that don't have any values. Did something go wrong with the data export?

It's possible that those datasets contain variables that have no data. One possible reason for this is that the variables concerned have simply not been asked yet (e.g. due to the skip patterns implemented within the survey or to simply not being used at all). Please check the paper version of the survey available on the Survey page to find out if this is the case. Another potential reason is that these variables were merely used to present information to the respondent rather than soliciting an answer to a question. Should none of the above apply, then please contact us at alp@rand.org with details about the survey dataset (or set of variables from different surveys) that you were trying to download.

Do you weight collected data and if so, how?

Weights have been created for each survey and are available on request. Details on the weighting process can be found here. To request weights for a particular survey, please contact us at alp@rand.org

There is a lot of information available. How can I look for something specific?

Searches can be performed using the search box located in the top right. To conduct a search just enter a keyword, subject, or question number and hit enter. Search results are presented by category, classified either as web site related results, paper related results, or data related results.

Can I find out how questions in a survey were asked?

Under the "Metadata" header on each Survey Page you can download the survey paper version. This document describes which questions were asked in the survey and in accordance with which logic.

Is there a way to learn details about how many respondents participated in a survey?

Details on the response rate for individual surveys are accessible through the "response overview" link on each Survey Page. Information on how response rates are calculated can be found here.

Can I find out which respondents did not participate in a survey?

Demographics describing the non-respondents of a survey are available for all surveys from Well Being 14 and up. They can be downloaded on the Survey Page of each individual survey via the 'Non-respondent demographics' link. Each non-respondents dataset contains the demographics as they would have been for the non-participatory panel members had they participated in the survey. Alternatively, on the Question Page of every demographics question (located within the Demographics module) data aggregates can be viewed for non-respondents by clicking on the 'compare to non-respondents' link located beneath the data overview for respondents. NOTE: for surveys that are in the field the non-respondent data will change as more respondents participate.

How do you find the people who are participating in the American Life Panel surveys?

The American Life Panel recruits participants from several sources, including representative samples of U.S. consumers. Details on panel composition and recruitment methods can be found here.

How can I identify and match panel members that participated in multiple surveys?

Each American Life Panel member is assigned an unique identifier. This identifier is used during survey data collection to link data to particular members. Identifiers are of the form 12345678:1, where the part before the ":" (i.e. 12345678) represents the household identifier, whereas the suffix behind the ":" represent the household member identifier.

I have cases in my dataset where the identifier ends in ':2' or ':3'. What does that mean?

Although the American Life Panel is primarily an individual panel, there are households that consist of multiple respondents. In such cases all respondents within the household share the same household identifier (everything before the ":"), but differ in their suffix behind the ":". To find out the relationship between the primary respondent in the household (with suffix ":1") and other household members, you can consult the household members module in the Household information survey.

What happens if people move out of a multi-respondent household?

In the event that a household member moves out of a household consisting of multiple respondent but remains a panel member, s/he is assigned a new identifier. In order to be able to trace respondents in the data as they migrate in and/or out of households, each dataset for a survey contains a variable called 'oldprim_key' (located in the Demographics module of the survey), which will contain the previous identifier of a respondent. The variable will have no value for a respondent if s/he did not previously have a different identifier. A complete household migration file is available for download from the Survey Page of the household information survey.