OMG! Excessive Texting Tied to Risky Teen Behaviors



Similar documents
1. C. The formula for the confidence interval for a population mean is: x t, which was

Inference on Proportion. Chapter 8 Tests of Statistical Hypotheses. Sampling Distribution of Sample Proportion. Confidence Interval

Hypothesis testing. Null and alternative hypotheses

One-sample test of proportions

Practice Problems for Test 3

Center, Spread, and Shape in Inference: Claims, Caveats, and Insights

Z-TEST / Z-STATISTIC: used to test hypotheses about. µ when the population standard deviation is unknown

Confidence Intervals

Confidence Intervals. CI for a population mean (σ is known and n > 30 or the variable is normally distributed in the.

Lesson 17 Pearson s Correlation Coefficient

STA 2023 Practice Questions Exam 2 Chapter 7- sec 9.2. Case parameter estimator standard error Estimate of standard error

Exam 3. Instructor: Cynthia Rudin TA: Dimitrios Bisias. November 22, 2011

Determining the sample size

5: Introduction to Estimation

Math C067 Sampling Distributions

1 Correlation and Regression Analysis

Case Study. Normal and t Distributions. Density Plot. Normal Distributions

Chapter 7: Confidence Interval and Sample Size

% 60% 70% 80% 90% 95% 96% 98% 99% 99.5% 99.8% 99.9%

Chapter 14 Nonparametric Statistics

Mann-Whitney U 2 Sample Test (a.k.a. Wilcoxon Rank Sum Test)

Confidence Intervals for One Mean

Output Analysis (2, Chapters 10 &11 Law)

I. Chi-squared Distributions

Overview. Learning Objectives. Point Estimate. Estimation. Estimating the Value of a Parameter Using Confidence Intervals

Professional Networking

MEI Structured Mathematics. Module Summary Sheets. Statistics 2 (Version B: reference to new book)

Non-life insurance mathematics. Nils F. Haavardsson, University of Oslo and DNB Skadeforsikring

PSYCHOLOGICAL STATISTICS

Analyzing Longitudinal Data from Complex Surveys Using SUDAAN

The following example will help us understand The Sampling Distribution of the Mean. C1 C2 C3 C4 C5 50 miles 84 miles 38 miles 120 miles 48 miles

CHAPTER 7: Central Limit Theorem: CLT for Averages (Means)

University of California, Los Angeles Department of Statistics. Distributions related to the normal distribution


Lesson 15 ANOVA (analysis of variance)

This document contains a collection of formulas and constants useful for SPC chart construction. It assumes you are already familiar with SPC.

Unit 8: Inference for Proportions. Chapters 8 & 9 in IPS

Sampling Distribution And Central Limit Theorem

A Test of Normality. 1 n S 2 3. n 1. Now introduce two new statistics. The sample skewness is defined as:

AP Calculus AB 2006 Scoring Guidelines Form B

Confidence intervals and hypothesis tests

G r a d e. 2 M a t h e M a t i c s. statistics and Probability

Chapter 6: Variance, the law of large numbers and the Monte-Carlo method

Statistical inference: example 1. Inferential Statistics

Properties of MLE: consistency, asymptotic normality. Fisher information.

Chapter 7 Methods of Finding Estimators

Measures of Spread and Boxplots Discrete Math, Section 9.4

Topic 5: Confidence Intervals (Chapter 9)

Hypothesis testing using complex survey data

1 Computing the Standard Deviation of Sample Means

Research Method (I) --Knowledge on Sampling (Simple Random Sampling)

Maximum Likelihood Estimators.

STATISTICAL METHODS FOR BUSINESS

, a Wishart distribution with n -1 degrees of freedom and scale matrix.

Definition. A variable X that takes on values X 1, X 2, X 3,...X k with respective frequencies f 1, f 2, f 3,...f k has mean

Taking DCOP to the Real World: Efficient Complete Solutions for Distributed Multi-Event Scheduling

Best of security and convenience

Tradigms of Astundithi and Toyota

FM4 CREDIT AND BORROWING

Bond Valuation I. What is a bond? Cash Flows of A Typical Bond. Bond Valuation. Coupon Rate and Current Yield. Cash Flows of A Typical Bond

Chapter XIV: Fundamentals of Probability and Statistics *

Overview of some probability distributions.

France caters to innovative companies and offers the best research tax credit in Europe

Chapter 7 - Sampling Distributions. 1 Introduction. What is statistics? It consist of three major areas:

The Forgotten Middle. research readiness results. Executive Summary

SECTION 1.5 : SUMMATION NOTATION + WORK WITH SEQUENCES

Soving Recurrence Relations

3. If x and y are real numbers, what is the simplified radical form

Agenda. Outsourcing and Globalization in Software Development. Outsourcing. Outsourcing here to stay. Outsourcing Alternatives

Multi-server Optimal Bandwidth Monitoring for QoS based Multimedia Delivery Anup Basu, Irene Cheng and Yinzhe Yu

Section 11.3: The Integral Test

Normal Distribution.

GOOD PRACTICE CHECKLIST FOR INTERPRETERS WORKING WITH DOMESTIC VIOLENCE SITUATIONS

AP Calculus BC 2003 Scoring Guidelines Form B

Asymptotic Growth of Functions

GCSE STATISTICS. 4) How to calculate the range: The difference between the biggest number and the smallest number.

Modified Line Search Method for Global Optimization

1 The Gaussian channel

THE ARITHMETIC OF INTEGERS. - multiplication, exponentiation, division, addition, and subtraction

A Combined Continuous/Binary Genetic Algorithm for Microstrip Antenna Design

MARTINGALES AND A BASIC APPLICATION

A GUIDE TO LEVEL 3 VALUE ADDED IN 2013 SCHOOL AND COLLEGE PERFORMANCE TABLES

Week 3 Conditional probabilities, Bayes formula, WEEK 3 page 1 Expected value of a random variable

PROCEEDINGS OF THE YEREVAN STATE UNIVERSITY AN ALTERNATIVE MODEL FOR BONUS-MALUS SYSTEM

INVESTMENT PERFORMANCE COUNCIL (IPC)

THE ROLE OF EXPORTS IN ECONOMIC GROWTH WITH REFERENCE TO ETHIOPIAN COUNTRY

LECTURE 13: Cross-validation

The Stable Marriage Problem

PFF2 2015/16. Assessment of Financial Circumstances For parents and partners of students. /SFEngland. /SF_England SFE/PFF2/1516/B

Investing in Stocks WHAT ARE THE DIFFERENT CLASSIFICATIONS OF STOCKS? WHY INVEST IN STOCKS? CAN YOU LOSE MONEY?

Characterizing End-to-End Packet Delay and Loss in the Internet

Now here is the important step

Is there employment discrimination against the disabled? Melanie K Jones i. University of Wales, Swansea

National Institute on Aging. What Is A Nursing Home?

Tell us if you need help because of a disability Ask for a free interpreter

Building Blocks Problem Related to Harmonic Series

CHAPTER 3 THE TIME VALUE OF MONEY

In nite Sequences. Dr. Philippe B. Laval Kennesaw State University. October 9, 2008

UM USER SATISFACTION SURVEY Final Report. September 2, Prepared by. ers e-research & Solutions (Macau)

Get advice now. Are you worried about your mortgage? New edition

Transcription:

BUSIESS WEEK: EXECUTIVE EALT ovember 09, 2010 OMG! Excessive Textig Tied to Risky Tee Behaviors Kids who sed more tha 120 a day more likely to try drugs, alcohol ad sex, researchers fid TUESDAY, ov. 9 (ealthday ews) -- Excessive textig ad social etworkig may icrease tees' risk for dagerous health behaviors, icludig smokig, drikig ad sexual activity, a ew study suggests. Researchers looked at hyper-textig (sedig more tha 120 messages per school day) ad hyper-etworkig (spedig more tha three hours a school day o social etworkig sites) amog high school studets i a urba couty i the U.S. Midwest. May of the 19.8 percet of tees who reported hyper-textig were female, miority, from lower socioecoomic status ad had o father at home, accordig to the researchers at Case Wester Reserve School of Medicie i Clevelad. yper-texters were: 40 percet more likely to have tried smokig; two times more likely to have tried alcohol; 43 percet more likely to bige-drik; 41 percet more likely to have used illicit drugs; 55 percet more likely to have bee i a physical fight; early 3.5 times more likely to have had sex; ad 90 percet more likely to have had four or more sexual parters. The 11.5 percet of studets who were hyper-etworkers were: 62 percet more likely to have smoked cigarettes; 79 percet more likely to have tried alcohol; 69 percet more likely to be bige drikers; 84 percet more likely to have used illicit drugs; 94 percet more likely to have bee i a physical fight; 69 percet more likely to have had sex; ad 60 percet more likely to have had four or more sexual parters. yper-etworkig was also associated with icreased likelihood of stress, depressio, suicide, poor sleep, poor academics, televisio watchig ad paretal permissiveess. The study was to be preseted Tuesday at the America Public ealth Associatio aual meetig i Dever. "The startlig results of this study suggest that whe left uchecked, textig ad other widely popular methods of stayig coected ca have dagerous health effects o teeagers," lead researcher Dr. Scott Frak, director of the School of Medicie's Master of Public ealth Program, said i a uiversity ews release. "This should be a wake-up call for parets to ot oly help their childre stay safe by ot textig ad drivig, but by discouragig excessive use of the cell phoe or social web sites i geeral," he added. SOURCE: Case Wester Reserve School of Medicie, ews release, ov. 9, 2010 Copyright 2010 ealthday. All rights reserved. OMG! Page 1

Uits: Teeagers (from urba areas i the Midwest we cofie coclusios to such teeagers). Explaatory Variable: Whether or ot a perso is a hypertexter. Categorical This is a observatioal study. 2 categories / levels Respose Variable: Whether or ot a perso has tried illicit drugs. Categorical 4 rows of the data table, showig all level combiatios Summary of the raw data two categories / levels Teeager Textig Activity Use of Illicit Drugs Illicit Drug Use Wada yper Tried Textig Activity Tried ot Tried Total Xavier yper ot tried yper 51 65 116 Yolada ohyper Tried ohyper 45 98 143 Zach ohyper ot tried Totals 96 163 259 ere s a clear, simple, ad effective report of the survey results. % of teeagers who have tried illicit drugs Amog hypertexters: 44.0% ( 116) Amog ohypertexters: 31.5% ( 143) 95% cofidece itervals for the proportios of all teeagers who have tried illicit drugs: Amog hypertexters: 0.4397 0.0903 0.3494 < p < 0.5300 Amog ohypertexters: 0.3147 0.0761 0.2386 < p < 0.3908 Goal: A 95% cofidece iterval for the differece betwee populatio proportios, p p. With a categorical (two levels) respose variable ad categorical (two levels) explaatory variable we hope to use the two sample Z procedures for the differece betwee proportios. Requiremet: Radom samplig; Idepedet samples; Populatio size at least 20 times the sample size; All couts i the summary of raw data at least 5. Poit estimate of differece: Error margi for differece: 95% cofidece iterval: < p p < Iterpretate the iterval: I am 95% cofidet that For a test ( 0.05) of 0 : p p 1 : p > p what s the decisio? OMG! Page 2

ere are the results of the hypertextig/illicit drugs iformatio, ow split by the how may parets variable. For teeagers with two parets 2 PARETS Illicit Drugs For the 2-paret teeagers, test Textig Tried ot Total ypertexters 17 55 72 ohypertexters 27 93 120 Total 44 138 192 Estimated differece Test statistic: Z p p P-value Coclusio: At the 5% level 0 : p p 1 : p p Pooled proportio p For teeagers with oe paret 1 PARET Illicit Drugs For the 1-paret teeagers, test Textig Tried ot Total ypertexters 34 10 44 ohypertexters 18 5 23 Total 52 15 67 Pooled proportio p Test statistic: Z p p P-value Coclusio: At the 5% level 0 : p p 1 : p p Estimated differece OMG! Page 3

For all teeagers ere are the results aggregated over the how may parets variable (as o page 1 we earlier obtaied the cofidece iterval). This assesses the hypertextig explaatory variable. ALL Illicit Drugs 0 : p p 1 : p p Textig Tried ot Total Test statistic: Z 2.07 ypertexters 51 65 116 P-value 0.019 ohypertexters 45 98 143 95% cofidece iterval: Total 96 163 259 0.007 < p p < 0.273 Coclusio: At the 5% level Reread Dr. Frak s statemet. What do you thik? ere are the results aggregated over the whether or ot a hypertexter variable, ad istead compared o umber of parets (assessig the umber of parets explaatory variable). 1 The subscriptig o the p s is chaged to reflect this. otice that the totals match those from above ad o the other side. ALL Illicit Drugs 0 : p1 p2 1 : p1 p2 # of parets Tried ot Total ˆp 1 0.7761 ˆp 2 0.2292 1 52 15 67 2 44 148 192 Total 96 163 259 99% cofidece iterval: E 0.22920.7708 0.7761 0.2239 2.576 0.1527 67 192 p 0.5469 1 ˆ2 Z 7.98 P-value 99% CI: < p1 p2 < Coclusio: 1 This aalysis is appropriate, because we ca see from the split o the reverse side that there is virtually o effect of hypertextig o the likelihood of tryig illicit drug use. OMG! Page 4

Solutios Page 2 The poit estimate is 0.4397 0.3147 0.1250 2 2 The error margi is 0.0903 0.0761 0. 1181. This ca also be computed from scratch (without kowig the two idividual error margis): 0.4397 0.5603 0.3147 0.6853 116 143 0.1181 The cofidece iterval bouds are 0.125 ± 0.118 which gives 0.007 < p p < 0.273. I am 95% cofidet that the proportio of all teeage hypertexters who have tried illicit drugs is betwee 0.007 ad 0.273 higher tha the proportio of all ohypertexters who have tried illicit drugs. Sice the cofidece iterval implies that p is higher tha p, the decisio is to reject the ull hypothesis. Page 3 Teeagers with two parets The two estimated proportios are p ˆ 17 72 0. 2361 ad p ˆ 27 120 0. 2250. The estimate differece is 0.0111. (ot very big at all.) The pooled proportio is p 17 27 72 120 44 192 0.. 2292 The test statistic is Z 0.2361 0.2250 0 0.2292 0.7708 0.2292 0.7708 72 120 0.0111 0.29. 0.626 This is a right tailed test; the P-value is the area to the right of 0.29: P-value 0.3859. There is virtually o evidece i favor of the alterative hypothesis. At the 5% level there is to sufficiet evidece i the data to coclude that the proportio of all hypertexters who have tried illicit drugs is higher tha the proportio of all ohypertexters who have tried illicit drugs. Teeagers with oe paret The two estimated proportios are 0.7727 ad 0.7826. The estimated differece is -0.0099 that s about 1% ad i the wrog directio from what is suggested i the alterative hypothesis. There is O evidece agaist the ull here. The pooled proportio is 0.7761. The deomiator of the test statistic (the stadard error SE ) is 0.0879. The test statistic is -0.0099/0.0879-0.11. The P-value is 0.5438. At the 5% level there is isufficiet evidece i the data to coclude that the proportio of all hypertexters who have tried illicit drugs is higher tha the proportio of all ohypertexters who have tried illicit drugs. I fact: There is O evidece of this. The observed differece is opposite i sig to this. OMG! Page 5

Page 4 Aggregated over the how may parets variable, assessig the effects of hypertextig At the 5% level there is sufficiet evidece i the sample data to coclude that hypertextig tees are more likely to have tried illicit drugs tha are ohypertextig tees. We re 95% cofidet that the differece i proportios (percets) is betwee 0.007 (0.7%) ad 0.273 (27.3%). owever: That does t mea at all that textig behavior is a cause of illicit drug use. As we see above, it may well be that the root cause is how may parets a teeager has. There certaily is more i favor of such a explaatio. Aggregated over the whether or ot a hypertexter variable, assessig the effects of paretal presece. The two observed proportios are 0.7761 ad 0.2292. Their differece is 0.5469. The P-value for the test is essetially 0. The 99% CI has bouds give by 0.5469 ± 0.1527, which gives 0.3992 < p1 p2 < 0.6996. We are 99% cofidet that the proportio of all tees from 1-paret families who have tried illicit drugs is betwee 0.3942 ad 0.6996 higher tha that for tees from 2-paret families. (Betwee roughly 40% ad 70% higher.) owever: This too may ot be iterpreted as causal while the umber of parets variable is certaily more strogly associated with drug use, this does ot imply that the umber of parets is a cause of illicit drug use. What about icome? I a similar way as show above, we might well fid that tees from lower icome families are simultaeously more like to have 1 paret ad to have tried illicit drugs. I short: It is oe thig to establish that two variables are associated. It is aother altogether to establish that chages i a explaatory variable cause a chage i the distributio of the respose variable. From a sigle observatioal study, oe ca ever coclude o the issue of causatio. OMG! Page 6