netrashetty
Netra Shetty
Energizer Holdings (NYSE: ENR) is an American manufacturer of batteries and personal care products, headquartered in Town and Country, Missouri.[2][3][4][5] Its most well known brands are Energizer and Eveready batteries, Schick, Wilkinson Sword and Edge shaving products, Playtex feminine hygiene and baby products, and Hawaiian Tropic and Banana Boat sunscreen products. The company sells in over 165 countries
Traditionally, surveys have been used to elicit unobservable information about peopleís valuations of goods when markets and prices for those goods are absent. They can also be
a valuable source of information when markets exist. People outside a market can assess
descriptions about the items being sold in that market, and this paper shows how surveys
can be used to exploit their assessments to generate a measure of the amount of information
and/or heterogeneity of preferences within that market. SpeciÖcally, in markets where participants receive di§erent signals about an itemís value due to noise and/or due to di§erent
costs and preferences, the survey can be used to estimate the characteristics (mean, variance)
of the distribution of those signals.
This measure can be employed as a regressor in empirical work where variance in the dependent variable (e.g., auction prices, retail price dispersion, or investment choices in stocks,
R&D, or education) might be explained by uncertainty about the value of the item being
sold or the returns to investment choice and/or heterogeneous preferences in the market.
The e§ects of incomplete information and heterogeneous preferences are usually relegated
to the error term, which a) confounds these e§ects with other drivers of the error term and
b) could lead to heteroskedasticity at best or omitted variable bias at worst. Furthermore,
by speciÖcally modeling the e§ect of this uncertainty or dispersed taste, one can estimate
policy implications such as the e§ect of publicly introducing information into the market or
selecting the pool of agents to change the distribution of preferences (Yin 2005).
Consider an empirical setting that examines the di§erence between prices and choices
over a variety of goods. Examples include the dispersion of retail prices over di§erent types
of drugs (Sorensen 2000), the choice (or level) of investment in di§erent types of Önancial
instruments, the R&D contributions in di§erent industries, the choice to attain di§erent
levels of education, or the prices of di§erent items being auctioned. Survey responses could
be used to produce a measure of the dispersion of information regarding the e¢ cacy of the
drug or the returns to di§erent types of investments or levels of education. In the auction
setting, including the variance of survey responses in a price regression allows one to control
for the dispersion of information in the market if the item being auctioned has a common
value or control for the dispersion of preferences in the market if the item being auctioned
has a private value. In all these examples, the regression model is misspeciÖed if one fails
to account for di§ering preferences or information. At best, this leads to heteroskedasticity, since the di§erent goods all have di§erent error distributions arising from variance in
information or preferences. At worst, this omitted variable will bias coe¢ cient estimates of
the other regressors if those regressors are correlated with the realization of preferences or
information. In an auction setting, survey data has the added beneÖt of being measured
independently from data generated during the auction itself. By deÖnition, data from the
auction is a function of bidder behavior. This makes the external survey data useful for testing hypotheses about bidding behavior in the auction that otherwise could not be conducted
without making assumptions about the nature of the private information signals.
Researchers often avoid using surveys due to the time and e§ort involved in conducting
2them. However, use of online surveys reduces some of the cost. This paper suggests a
survey design technique and econometric tool to deal with a general population of survey
respondents, including those who participate in the market of interest and those who do not.
The use of inexperienced respondents permits the survey to be implemented more quickly
and with a larger number of respondents than if the researcher had to restrict her search
for an equivalent number of experienced respondents. The use of experienced respondents
allows me to correct for potential bias from using more noisy, inexperienced responses.
To the extent that the survey is still more costly to conduct than gathering observable
data, this paper argues that the survey data is more valuable because it exploits the human
ability to assess complex information sets in a way that cannot be accomplished by hedonic
evaluation. Often, hedonics are used to control for the value of the good. However, hedonic
methods su§er from the need to deÖne a good into a limited set of characteristics, and it
does not provide any means for taking into account anomalies in products that may not
Öt any category. Survey data exploits human assessment of information to collapse many
dimensions into a single numerical value. This does not preclude the econometrician from
also employing hedonic measures along side the survey data.
The particular application used here is for eBay online auctions for personal computers
(PCs). In all auctions, private information signals (not directly observable to the econometrician) about the value of the item being sold is dispersed among the auction participants.
I used a survey to measure the mean and dispersion of those information signals in computer
auctions.
In a common values (CV) auction setting, each auction participantís private signal contains information that is relevant to the other participantsíassessments of the value of the
item. In this setting, the average of these survey responses provides a potential measure of
the common value of the item being auctioned. The standard deviation of responses provides
a potential measure of the dispersion of information in the auction. An auction where more
information is publicly available to all the bidders will be reáected in less dispersed signals.
In a private values (PV) setting, a private signal only concerns the recipientís own value
for the item. In this case, the survey measures the average private value and dispersion of
private values among bidders. One can use these averages and standard deviations to test
between PV and CV settings (or the dominant component if the setting is mixed) while also
testing for Bayesian-Nash equilibrium bidding behavior (see Yin 2005).
Analysis of the survey results conÖrms that the survey is able to successfully generate
estimates of information dispersion and average item values. Auction descriptions matched
results generated in the survey. Auctions which my survey respondents designated to be
of equal value contained equivalent hardware speciÖcations. The auction description that
provided more details (i.e., revealed more information to all auction participants) had a lower
standard deviation of survey respondentís valuations. The price attained in that auction
was higher than that attained for the item with a less informative auction description. This
last Önding is consistent with the auction theory prediction that prices decline with more
information dispersion in CV settings.
Four in Five Spend Their Own Money Purchasing Supplies or Equipment for their Department
Two thirds (64%) of volunteer firefighters report that their department does not have sufficient funds to sustain all areas in which it operates, according to a recent Ipsos poll conducted on behalf of Duracell, a leading manufacturer of high-performing alkaline batteries.
The situation seems particularly critical in the South, where 70% of volunteer firefighters believe their departments do not receive enough funds. In addition, the survey shows that fire chiefs, are also more likely than other volunteer firefighters to believe their department is underfunded (80%).
Using their Own Money
The survey shows that over four in five (86%) volunteer firefighters dip into their own pockets to purchase supplies and/or equipment for their department. The numbers are spread almost evenly across the Northeast (84%), Midwest (84%), South (89%) and West (87%).
Of those who report using their own money, eight in ten (80%) dispense over $100 a year; including more than half (55%) who say they spend $101 to $500, 17% who spend $501 to $1,000, and 8% who spend more than $1,000 a year.
Overall, volunteer firefighters in the Northeast (73%) are less likely than those in the South (86%) and the West (86%) to report they spend over $100 of their own money.
Volunteer Firefighters While Working Full-Time
In addition to spending their own money, most volunteer firefighters also spend a lot of their time in the fire department. Six in ten (59%) report to volunteer over 10 hours a week, including 24% who volunteer over 20 hours a week.
However, when they are not serving their community, most volunteer firefighters also have a full-time job; with 63% of them reporting that in addition to being a firefighter, they work over 40 hours a week in their paid profession. An additional 21% report to work 21 to 40 hours a week, while just a small proportion are either retired (10%) or work under 20 hours a week (3%).
Tough Hours
Most volunteer firefighters say they are most often call into action in the evening and wee hours of the morning; as 64% of them report that the most common time of the day they are called upon to serve is between 6 p.m. and 6 a.m. – this includes half of all respondents (51%) who report to be called on duty between 6 p.m. and midnight, and 13% who report they are most often called between midnight and 6 a.m.
In contrast, only a quarter of all respondents (25%) report being called to serve most often during the day, including 19% who say they are most frequently called between noon and 6.p.m., and just 6% who report being called during the 6 a.m. and noon.
These are some of the findings of an online Ipsos poll conducted January 4 - 11, 2011. For the survey, a national sample of 533 volunteer firefighters from the National Volunteer Fire Council was interviewed online.
A probability sample of 533 respondents, with a 100% response rate, would have an estimated margin of error of +/-4.2 percentage points, 19 times out of 20 of what the results would have been had the entire population of volunteer firefighters in the United States been polled.
All sample surveys and polls may be subject to other sources of error, including, but not limited to coverage error, and measurement error.
Traditionally, surveys have been used to elicit unobservable information about peopleís valuations of goods when markets and prices for those goods are absent. They can also be
a valuable source of information when markets exist. People outside a market can assess
descriptions about the items being sold in that market, and this paper shows how surveys
can be used to exploit their assessments to generate a measure of the amount of information
and/or heterogeneity of preferences within that market. SpeciÖcally, in markets where participants receive di§erent signals about an itemís value due to noise and/or due to di§erent
costs and preferences, the survey can be used to estimate the characteristics (mean, variance)
of the distribution of those signals.
This measure can be employed as a regressor in empirical work where variance in the dependent variable (e.g., auction prices, retail price dispersion, or investment choices in stocks,
R&D, or education) might be explained by uncertainty about the value of the item being
sold or the returns to investment choice and/or heterogeneous preferences in the market.
The e§ects of incomplete information and heterogeneous preferences are usually relegated
to the error term, which a) confounds these e§ects with other drivers of the error term and
b) could lead to heteroskedasticity at best or omitted variable bias at worst. Furthermore,
by speciÖcally modeling the e§ect of this uncertainty or dispersed taste, one can estimate
policy implications such as the e§ect of publicly introducing information into the market or
selecting the pool of agents to change the distribution of preferences (Yin 2005).
Consider an empirical setting that examines the di§erence between prices and choices
over a variety of goods. Examples include the dispersion of retail prices over di§erent types
of drugs (Sorensen 2000), the choice (or level) of investment in di§erent types of Önancial
instruments, the R&D contributions in di§erent industries, the choice to attain di§erent
levels of education, or the prices of di§erent items being auctioned. Survey responses could
be used to produce a measure of the dispersion of information regarding the e¢ cacy of the
drug or the returns to di§erent types of investments or levels of education. In the auction
setting, including the variance of survey responses in a price regression allows one to control
for the dispersion of information in the market if the item being auctioned has a common
value or control for the dispersion of preferences in the market if the item being auctioned
has a private value. In all these examples, the regression model is misspeciÖed if one fails
to account for di§ering preferences or information. At best, this leads to heteroskedasticity, since the di§erent goods all have di§erent error distributions arising from variance in
information or preferences. At worst, this omitted variable will bias coe¢ cient estimates of
the other regressors if those regressors are correlated with the realization of preferences or
information. In an auction setting, survey data has the added beneÖt of being measured
independently from data generated during the auction itself. By deÖnition, data from the
auction is a function of bidder behavior. This makes the external survey data useful for testing hypotheses about bidding behavior in the auction that otherwise could not be conducted
without making assumptions about the nature of the private information signals.
Researchers often avoid using surveys due to the time and e§ort involved in conducting
2them. However, use of online surveys reduces some of the cost. This paper suggests a
survey design technique and econometric tool to deal with a general population of survey
respondents, including those who participate in the market of interest and those who do not.
The use of inexperienced respondents permits the survey to be implemented more quickly
and with a larger number of respondents than if the researcher had to restrict her search
for an equivalent number of experienced respondents. The use of experienced respondents
allows me to correct for potential bias from using more noisy, inexperienced responses.
To the extent that the survey is still more costly to conduct than gathering observable
data, this paper argues that the survey data is more valuable because it exploits the human
ability to assess complex information sets in a way that cannot be accomplished by hedonic
evaluation. Often, hedonics are used to control for the value of the good. However, hedonic
methods su§er from the need to deÖne a good into a limited set of characteristics, and it
does not provide any means for taking into account anomalies in products that may not
Öt any category. Survey data exploits human assessment of information to collapse many
dimensions into a single numerical value. This does not preclude the econometrician from
also employing hedonic measures along side the survey data.
The particular application used here is for eBay online auctions for personal computers
(PCs). In all auctions, private information signals (not directly observable to the econometrician) about the value of the item being sold is dispersed among the auction participants.
I used a survey to measure the mean and dispersion of those information signals in computer
auctions.
In a common values (CV) auction setting, each auction participantís private signal contains information that is relevant to the other participantsíassessments of the value of the
item. In this setting, the average of these survey responses provides a potential measure of
the common value of the item being auctioned. The standard deviation of responses provides
a potential measure of the dispersion of information in the auction. An auction where more
information is publicly available to all the bidders will be reáected in less dispersed signals.
In a private values (PV) setting, a private signal only concerns the recipientís own value
for the item. In this case, the survey measures the average private value and dispersion of
private values among bidders. One can use these averages and standard deviations to test
between PV and CV settings (or the dominant component if the setting is mixed) while also
testing for Bayesian-Nash equilibrium bidding behavior (see Yin 2005).
Analysis of the survey results conÖrms that the survey is able to successfully generate
estimates of information dispersion and average item values. Auction descriptions matched
results generated in the survey. Auctions which my survey respondents designated to be
of equal value contained equivalent hardware speciÖcations. The auction description that
provided more details (i.e., revealed more information to all auction participants) had a lower
standard deviation of survey respondentís valuations. The price attained in that auction
was higher than that attained for the item with a less informative auction description. This
last Önding is consistent with the auction theory prediction that prices decline with more
information dispersion in CV settings.
Four in Five Spend Their Own Money Purchasing Supplies or Equipment for their Department
Two thirds (64%) of volunteer firefighters report that their department does not have sufficient funds to sustain all areas in which it operates, according to a recent Ipsos poll conducted on behalf of Duracell, a leading manufacturer of high-performing alkaline batteries.
The situation seems particularly critical in the South, where 70% of volunteer firefighters believe their departments do not receive enough funds. In addition, the survey shows that fire chiefs, are also more likely than other volunteer firefighters to believe their department is underfunded (80%).
Using their Own Money
The survey shows that over four in five (86%) volunteer firefighters dip into their own pockets to purchase supplies and/or equipment for their department. The numbers are spread almost evenly across the Northeast (84%), Midwest (84%), South (89%) and West (87%).
Of those who report using their own money, eight in ten (80%) dispense over $100 a year; including more than half (55%) who say they spend $101 to $500, 17% who spend $501 to $1,000, and 8% who spend more than $1,000 a year.
Overall, volunteer firefighters in the Northeast (73%) are less likely than those in the South (86%) and the West (86%) to report they spend over $100 of their own money.
Volunteer Firefighters While Working Full-Time
In addition to spending their own money, most volunteer firefighters also spend a lot of their time in the fire department. Six in ten (59%) report to volunteer over 10 hours a week, including 24% who volunteer over 20 hours a week.
However, when they are not serving their community, most volunteer firefighters also have a full-time job; with 63% of them reporting that in addition to being a firefighter, they work over 40 hours a week in their paid profession. An additional 21% report to work 21 to 40 hours a week, while just a small proportion are either retired (10%) or work under 20 hours a week (3%).
Tough Hours
Most volunteer firefighters say they are most often call into action in the evening and wee hours of the morning; as 64% of them report that the most common time of the day they are called upon to serve is between 6 p.m. and 6 a.m. – this includes half of all respondents (51%) who report to be called on duty between 6 p.m. and midnight, and 13% who report they are most often called between midnight and 6 a.m.
In contrast, only a quarter of all respondents (25%) report being called to serve most often during the day, including 19% who say they are most frequently called between noon and 6.p.m., and just 6% who report being called during the 6 a.m. and noon.
These are some of the findings of an online Ipsos poll conducted January 4 - 11, 2011. For the survey, a national sample of 533 volunteer firefighters from the National Volunteer Fire Council was interviewed online.
A probability sample of 533 respondents, with a 100% response rate, would have an estimated margin of error of +/-4.2 percentage points, 19 times out of 20 of what the results would have been had the entire population of volunteer firefighters in the United States been polled.
All sample surveys and polls may be subject to other sources of error, including, but not limited to coverage error, and measurement error.
Last edited: