MAY 31, 2012 Twitter Use 2012 Overall adoption remains steady, but “typical day” usage continues to grow—8% of online adults now use Twitter on a typical day. African-Americans, young adults, and mobile users stand out for their high rates of Twitter usage. Aaron Smith Senior Research Specialist, Pew Internet Project Joanna Brenner Web Coordinator, Pew Internet Project Pew Research Center’s Internet & American Life Project 1615 L St., NW – Suite 700 Washington, . 20036 Phone: 202-419-4500
Findings Some 15% of online adults use Twitter as of February 2012, and 8% do so on a typical day. Although overall Twitter usage has nearly doubled since the Pew Research Center’s Internet & American Life Project first asked a stand-alone Twitter question in November 2010, the 15% of online adults who use Twitter as of early 2012 is similar to the 13% of such adults who did so in May 2011. At the same time, the proportion of online adults who use Twitter on a typical day has doubled since May 2011 and has quadrupled since late 2010—at that point just 2% of online adults used Twitter on a typical day1. The rise of smartphones might account for some of the uptick in usage because smartphone users are particularly likely to be using Twitter. Twitter usage over time % of internet users who use Twitter 40% 20% 15% 13% 12% 8% % 0%2%4% 5%8 November 2010 May 2011 August 2011 February 2012 Typical day Total Source: Pew Research Center's Internet & American Life Project Winter 2012 Tracking Survey, January 20-February 19, 2012. N=2,253 adults age 18 and older, including 901 cell phone interviews. Interviews conducted in English and Spanish. Margin of error is +/ percentage points for internet users (n=1,729). Several demographic groups stand out as having high rates of Twitter usage relative to their peers: ξAfrican-Americans — Black internet users continue to use Twitter at high rates. More than one quarter of online African-Americans (28%) use Twitter, with 13% doing so on a typical day. ξYoung adults — One quarter (26%) of internet users ages 18-29 use Twitter, nearly double the rate for those ages 30-49. Among the youngest internet users (those ages 18-24), fully 31% are Twitter users. 1 We describe activities on a “typical day” by using data we gather from respondents in answer to questions we ask them about what they did “yesterday”—that is, the day before we reached them for our survey. 2
ξUrban and suburban residents — Residents of urban and suburban areas are significantly more likely to use Twitter than their rural counterparts. Who uses Twitter? % of internet users within each group who use Twitter All adult internet users (n=1729) 15% Men (n=804) 14 Women (n=925) 15 Age 18-29 (n=316) 26** 30-49 (n=532) 14 50-64 (n=521) 9 65+ (n=320) 4 Race/ethnicity White, Non-Hispanic (n=1229) 12 Black, Non-Hispanic (n=172) 28** Hispanic (n=184) 14 Annual household income Less than $30,000/yr (n=390) 19 $30,000-$49,999 (n=290) 12 $50,000-$74,999 (n=250) 14 $75,000+ (n=523) 17 Education level No high school diploma2 (n=108) 22 High school grad (n=465) 12 Some College (n=447) 14 College + (n=698) 17 Geographic location Urban (n=520) 19** Suburban (n=842) 14** Rural (n=280) 8 Source: Pew Research Center's Internet & American Life Project Winter 2012 Tracking Survey, January 20-February 19, 2012. N=2,253 adults age 18 and older, including 901 cell phone interviews. Interviews conducted in English and Spanish. The margin of error is +/ percentage points for internet users. **Represents significant difference compared with all other rows in group. 2 Due to a small number of respondents in this group in our May 2011 survey, we did not report individually on the “no high school diploma” group in our 2011 report on Twitter usage. 3
Twitter use among 18-24 year olds increased dramatically between May 2011 and February 2012, both overall and on a “typical day” basis Twitter use within the overall population remained steady over the last year, and usage rates within most major demographic groups changed little over the same time period. The youngest adults (those between the ages of 18 and 24) are the primary exception to this trend—nearly one third of internet users in this age group now use Twitter, up from 18% in May of 2011 and 16% in late 20103. Twitter use by those in their mid-20s to mid-40s largely leveled off in the last year after roughly doubling between late 2010 and mid 2011. Twitter adoption by age, 2010-2012 % of internet users in each group who use Twitter November 2010 May 2011 February 2012 All adults 8% 13% 15% 18-24 16 18 31 25-34 9 19 17 35-44 8 14 16 45-54 7 9 9 55-64 4 8 9 65+ 4 6 4 Sources: Pew Research Center’s Internet & American Life Project tracking surveys. 2012 data based on January 20-February 19, 2012 Tracking Survey. N=2,253 adults age 18 and older, including 901 cell phone interviews, margin of error is +/ percentage points based on internet users (n=1729). In addition to increasing on an overall basis, the proportion of young internet users who use Twitter on a typical day also doubled over the last year. Fully one in five internet users ages 18-24 (20%) now use Twitter on a typical day, up from 9% in May 2011. Notably, “typical day” usage among slightly older adults (those ages 25-34) also doubled—from 5% of such internet users in May 2011 to 11% in February 2012—even as overall usage levels within this group remained stable over that time period. 3 This increase among 18-24 year olds did not manifest itself in higher usage rates within the overall population because this group is relatively small; 18-24 year olds represent 16% of the adult internet user population. 4
“Typical day” Twitter use by age, 2010-2012 % of internet users in each group who use Twitter on a typical day November 2010 May 2011 February 2012 All adults 2% 4% 8% 18-24 4 9 20 25-34 5 5 11 35-44 2 6 9 45-54 2 3 3 55-64 1 2 4 65+ <1 <1 1 Sources: Pew Research Center’s Internet & American Life Project tracking surveys. 2012 data based on January 20-February 19, 2012 Tracking Survey. N=2,253 adults age 18 and older, including 901 cell phone interviews, margin of error is +/ percentage points based on internet users (n=1729). Twitter and the “Mobile Difference” We can also see this relationship between youth, mobility and Twitter use when looking specifically at Twitter use on mobile phones. Twitter usage is highly correlated with the use of mobile technologies, especially smartphones. One in five smartphone owners (20%) are Twitter users, with 13% using the service on a typical day. By contrast, internet users who own more basic mobile phones are roughly half as likely to use Twitter overall (9% do so), and just 3% of these more basic phone owners are “typical day” users. Indeed, this correlation between Twitter adoption and smartphone ownership may help to explain the recent growth in Twitter usage among young adults. Those ages 18-24 are not just the fastest growing group when it comes to Twitter adoption over the last year—they also experienced the largest increase in smartphone ownership of any demographic group over the same time period. In addition to asking internet users whether they ever use Twitter (regardless of the platform or device used) in our February 2012 tracking survey, we included a question in our April 2012 tracking survey in which we asked adult cell phone owners if they use Twitter specifically on their mobile phones. Overall we found that 9% of cell owners use Twitter on their phones, with 5% doing so on a typical As with general Twitter usage, smartphone owners are much more likely than average to use Twitter on their phones (overall 16% of smartphone owners use Twitter on their phones, and 10% do so on a typical day). 4 This question was asked of mobile phone owners who use the internet, email or apps on their phones; the results here have been recalculated based on all cell owners. 5
As with our general Twitter usage findings, cell owners ages 18-24 are more likely than older cell owners to use Twitter within the context of their mobile devices—fully one in five 18-24 year old cell owners (22%) use Twitter on their phones, and 15% do so on a typical day. African Americans and Latinos (both of whom have high rates of smartphone ownership) also stand out as heavy mobile Twitter users. Who uses Twitter on a cell phone? % of adult cell owners in each group who use Twitter on their phones All cell owners (n=1954) 9% Men (n=895) 9 Women (n=1059) 9 Age 18-24 (n=225) 22** 25-34 (n=230) 14 35-44 (n=276) 9 45-54 (n=371) 5 55-64 (n=387) 3 65+ (n=429) <1 Race/ethnicity White, Non-Hispanic (n=1404) 7 Black, Non-Hispanic (n=234) 17** Hispanic (n=180) 12** Annual household income Less than $30,000/yr (n=447) 7 $30,000-$49,999 (n=316) 12 $50,000-$74,999 (n=272) 11 $75,000+ (n=538) 9 Education level No high school diploma (n=156) 10 High school grad (n=542) 6 Some College (n=490) 9 College + (n=752) 11 Geographic location Urban (n=557) 10 Suburban (n=993) 9 Rural (n=316) 6 Source: Pew Research Center's Internet & American Life Project Spring 2012 Tracking Survey, March 15-April 3, 2012. N=2,254 adults age 18 and older, including 903 cell phone interviews. Interviews conducted in English and Spanish. Margin of error is +/ percentage points for cell owners. **Represents significant difference compared with all other rows in group. 6
Survey Questions and Methodology Winter Tracking Survey 2012 Final Topline 02/22/2012 Data for January 20–February 19, 2012 Princeton Survey Research Associates International for the Pew Research Center’s Internet & American Life Project Sample: n=2,253 national adults, age 18 and older, including 901 cell phone interviews Interviewing dates: – Margin of error is plus or minus 2 percentage points for results based on Total [n=2,253] Margin of error is plus or minus 3 percentage points for results based on internet users [n=1,729] WEB1 Next... Please tell me if you ever use the internet to do any of the following things. Do you ever use the internet to...[INSERT; RANDOMIZE]? / Did you happen to do this yesterday, or not? Based on all internet users [N=1,729] TOTAL HAVE ---------- EVER DONE DID HAVE NOT THIS YESTERDAY DONE THIS DON’T KNOW REFUSED Use Twitter Current 15 8 85 * 0 August 2011 12 5 88 * 0 May 2011 13 4 87 * 0 January 2011 10 n/a 90 * * December 2010 12 n/a 88 * 0 November 2010 8 2 92 0 * This report is based on the findings of a survey on Americans' use of the Internet. The results in this report are based on data from telephone interviews conducted by Princeton Survey Research Associates International from January 20 to February 19, 2012, among a sample of 2,253 adults, age 18 and older. Telephone interviews were conducted in English and Spanish by landline (1,352) and cell phone (901, including 440 without a landline phone). For results based on the total sample, one can say with 95% confidence that the error attributable to sampling is plus or minus percentage points. For results based Internet users (n=1,729), the margin of sampling error is plus or minus percentage points. In addition to sampling error, question wording and practical difficulties in conducting telephone surveys may introduce some error or bias into the findings of opinion polls. A combination of landline and cellular random digit dial (RDD) samples was used to represent all adults in the continental United States who have access to either a landline or cellular telephone. Both samples were provided by Survey Sampling International, LLC (SSI) according to PSRAI specifications. Numbers for the landline sample were selected with probabilities in proportion to their share of listed telephone households from active blocks (area code + exchange + two-digit block number) that contained three or more residential directory listings. The cellular sample was not list-assisted, but was drawn through a 7
systematic sampling from dedicated wireless 100-blocks and shared service 100-blocks with no directory-listed landline numbers. New sample was released daily and was kept in the field for at least five days. The sample was released in replicates, which are representative subsamples of the larger population. This ensures that complete call procedures were followed for the entire sample. At least 7 attempts were made to complete an interview at a sampled telephone number. The calls were staggered over times of day and days of the week to maximize the chances of making contact with a potential respondent. Each number received at least one daytime call in an attempt to find someone available. For the landline sample, interviewers asked to speak with the youngest adult male or female currently at home based on a random rotation. If no male/female was available, interviewers asked to speak with the youngest adult of the other gender. For the cellular sample, interviews were conducted with the person who answered the phone. Interviewers verified that the person was an adult and in a safe place before administering the survey. Cellular sample respondents were offered a post-paid cash incentive for their participation. All interviews completed on any given day were considered to be the final sample for that day. Weighting is generally used in survey analysis to compensate for sample designs and patterns of non-response that might bias results. A two-stage weighting procedure was used to weight this dual-frame sample. The first-stage corrected for different probabilities of selection associated with the number of adults in each household and each respondent’s telephone usage patterns. This weighting also adjusts for the overlapping landline and cell sample frames and the relative sizes of each frame and each sample. The second stage of weighting balances sample demographics to population parameters. The sample is balanced to match national population parameters for sex, age, education, race, Hispanic origin, region (. Census definitions), population density, and telephone usage. The Hispanic origin was split out based on nativity; born and . born. The White, non-Hispanic subgroup is also balanced on age, education and region. The basic weighting parameters came from a special analysis of the Census Bureau’s 2011 Annual Social and Economic Supplement (ASEC) that included all households in the United States. The population density parameter was derived from Census 2000 data. The cell phone usage parameter came from an analysis of the July-December 2010 National Health Interview Survey. Following is the full disposition of all sampled telephone numbers: 8
Sample Disposition Landline Cell 33,732 22,499 Total Numbers Dialed 1,396 274 Non-residential 1,483 47 Computer/Fax 8 ---- Cell phone 14,936 8,237 Other not working 3,094 467 Additional projected not working 12,815 13,474 Working numbers % % Working Rate 1,031 156 No Answer / Busy 4,290 5,288 Voice Mail 40 16 Other Non-Contact 7,454 8,014 Contacted numbers % % Contact Rate 513 1,256 Callback 5,491 5,273 Refusal 1,450 1,485 Cooperating numbers % % Cooperation Rate 67 41 Language Barrier ---- 524 Child's cell phone 1,383 920 Eligible numbers % % Eligibility Rate 31 19 Break-off 1,352 901 Completes % % Completion Rate % % Response Rate The disposition reports all of the sampled telephone numbers ever dialed from the original telephone number samples. The response rate estimates the fraction of all eligible respondents in the sample that were ultimately interviewed. At PSRAI it is calculated by taking the product of three component rates: ξContact rate – the proportion of working numbers where a request for interview was made ξCooperation rate – the proportion of contacted numbers where a consent for interview was at least initially obtained, versus those refused ξCompletion rate – the proportion of initially cooperating and eligible interviews that were completed Thus the response rate for the landline sample was 11 percent. The response rate for the cellular sample was 11 percent. 9
Spring Tracking Survey 2012 Final Topline 04/10/2012 Data for March 15–April 3, 2012 Princeton Survey Research Associates International for the Pew Research Center’s Internet & American Life Project Sample: n=2,254 national adults, age 18 and older, including 903 cell phone interviews Interviewing dates: – Margin of error is plus or minus 2 percentage points for results based on Total [n=2,254] Margin of error is plus or minus 3 percentage points for results based on cell phone owners [n=1,954] Margin of error is plus or minus 4 percentage points for results based on those who use the internet or email on their cell phone or download apps to their cell phone [n=953] MOB1 Next... Please tell me if you ever use your cell phone to do any of the following things. Do you ever use your cell phone to... [INSERT ITEMS; RANDOMIZE]? / [IF YES: Did you happen to do this YESTERDAY, or not?] Based on those who use the internet or email on their cell phone or download apps to their cell phone [N=953] TOTAL HAVE ---------- EVER DONE DID HAVE NOT THIS YESTERDAY DONE THIS DON’T KNOW REFUSED Use Twitter Current 16 9 84 * 0 This report is based on the findings of a survey on Americans' use of the Internet. The results in this report are based on data from telephone interviews conducted by Princeton Survey Research Associates International from March 15 to April 3, 2012, among a sample of 2,254 adults, age 18 and older. Telephone interviews were conducted in English and Spanish by landline (1,351) and cell phone (903, including 410 without a landline phone). For results based on the total sample, one can say with 95% confidence that the error attributable to sampling is plus or minus percentage points. For results based Internet users (n=1,803), the margin of sampling error is plus or minus percentage points. In addition to sampling error, question wording and practical difficulties in conducting telephone surveys may introduce some error or bias into the findings of opinion polls. A combination of landline and cellular random digit dial (RDD) samples was used to represent all adults in the continental United States who have access to either a landline or cellular telephone. Both samples were provided by Survey Sampling International, LLC (SSI) according to PSRAI specifications. Numbers for the landline sample were selected with probabilities in proportion to their share of listed telephone households from active blocks (area code + exchange + two-digit block number) that contained three or more residential directory listings. The cellular sample was not list-assisted, but was drawn through a systematic sampling from dedicated wireless 100-blocks and shared service 100-blocks with no directory-listed landline numbers. New sample was released daily and was kept in the field for at least five days. The sample was released in replicates, which are representative subsamples of the larger population. This ensures that complete call procedures were followed for the entire sample. At least 7 attempts were made to complete an 10
interview at a sampled telephone number. The calls were staggered over times of day and days of the week to maximize the chances of making contact with a potential respondent. Each number received at least one daytime call in an attempt to find someone available. For the landline sample, interviewers asked to speak with the youngest adult male or female currently at home based on a random rotation. If no male/female was available, interviewers asked to speak with the youngest adult of the other gender. For the cellular sample, interviews were conducted with the person who answered the phone. Interviewers verified that the person was an adult and in a safe place before administering the survey. Cellular sample respondents were offered a post-paid cash incentive for their participation. All interviews completed on any given day were considered to be the final sample for that day. Weighting is generally used in survey analysis to compensate for sample designs and patterns of non-response that might bias results. A two-stage weighting procedure was used to weight this dual-frame sample. The first-stage corrected for different probabilities of selection associated with the number of adults in each household and each respondent’s telephone usage patterns. This weighting also adjusts for the overlapping landline and cell sample frames and the relative sizes of each frame and each sample. The second stage of weighting balances sample demographics to population parameters. The sample is balanced to match national population parameters for sex, age, education, race, Hispanic origin, region (. Census definitions), population density, and telephone usage. The Hispanic origin was split out based on nativity; born and . born. The White, non-Hispanic subgroup is also balanced on age, education and region. The basic weighting parameters came from a special analysis of the Census Bureau’s 2011 Annual Social and Economic Supplement (ASEC) that included all households in the United States. The population density parameter was derived from Census 2000 data. The cell phone usage parameter came from an analysis of the July-December 2010 National Health Interview Survey Following is the full disposition of all sampled telephone numbers: 11
Table 2:Sample Disposition Landline Cell 33,738 22,143 Total Numbers Dialed 1,502 332 Non-residential 1,491 45 Computer/Fax 8 ---- Cell phone 15,401 8,237 Other not working 2,746 404 Additional projected not working 12,590 13,126 Working numbers % % Working Rate 915 135 No Answer / Busy 3,472 4,465 Voice Mail 66 5 Other Non-Contact 8,137 8,521 Contacted numbers % % Contact Rate 523 1,382 Callback 6,161 5,654 Refusal 1,453 1,485 Cooperating numbers % % Cooperation Rate 52 43 Language Barrier ---- 498 Child's cell phone 1,401 944 Eligible numbers % % Eligibility Rate 50 41 Break-off 1,351 903 Completes % % Completion Rate % % Response Rate The disposition reports all of the sampled telephone numbers ever dialed from the original telephone number samples. The response rate estimates the fraction of all eligible respondents in the sample that were ultimately interviewed. At PSRAI it is calculated by taking the product of three component rates: ξContact rate – the proportion of working numbers where a request for interview was made ξCooperation rate – the proportion of contacted numbers where a consent for interview was at least initially obtained, versus those refused ξCompletion rate – the proportion of initially cooperating and eligible interviews that were completed Thus the response rate for the landline sample was 11 percent. The response rate for the cellular sample was 11 percent. 12
